From domestic to international film success:
The mediating role of festival performance, and the moderating role of
arthouse-mainstream congruence between films and festivals.
Carlijn Rosa Tolman (10002202) Supervisor: dr. J.J. Ebbers Second reader: A. Tomaselli
Date: 27-2-2017
Master Thesis Business Administration
Entrepreneurship and Management in the Creative Industries track Amsterdam Business School, University of Amsterdam
Statement of originality
This document is written by Carlijn Rosa Tolman, who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Table of contents
Abstract 5
1. Introduction 6
2. Theory 12
2.1. Market signals in the creative industries 12
2.2. Combinations of and interactions between market signals 15
2.3. Success in foreign markets: The liability of foreignness 18
2.4. Field configuring events as conduits of market signals 19
2.5. Film festivals as conduits of market signals 22
2.6. The importance of film festival status 24
2.7. The importance congruence 25
3. Data and method 28
3.1. Research setting and data collection 28
3.2. Dependent variables 28
3.3. Independent variables 29
3.4. Control variables 31
4. Results 33
4.1. Descriptive statistics and correlations 33
4.2. Hypothesis testing 35
Mediation 35
Moderated mediation 42
4.3. Robustness checks 44
5. Discussion 46
5.3. Managerial implications 50
5.4. Limitations and directions for future research 51
6. Conclusion 53
7. References 54
8. Appendix 58
Abstract
The quality of products in the creative industries can only be determined after consumption. Therefore, stakeholders rely on market signals. An important market signal in the creative industries is status. This research argues that being selected for a film festival, being nominated for an award at a festival or winning an award at a festival increases a movie’s status and, therefore, mediates the relationship between a movie’s performance in its domestic market and foreign performance. Moreover, the relationship between film festival performance and foreign performance is expected to be moderated by the film festival’s status and by the level of congruence between a movie’s and a festival’s position on the art house versus mainstream continuum. The results of the analysis do not provide support for this moderated mediation effect. However, the lack of support for the hypotheses could be caused by limitations of the data that were used and the question whether status and congruence influence the effectiveness of film festival performance as a market signal remains relevant.
1. Introduction
Once a year movie stars and other movie industry artists find all eyes directed at them, while
they stride down the red carpet in the most exquisite formal garb, to gather in a theatre in the
south of France and witness how little gold statuettes are handed out to those who showed
outstanding performance in the film industry over the past year. The Cannes Film Festival.
We have all heard of it and every year film studios spent large amounts of money on
advertising campaigns to increase their chances of being nominated for or even winning an
award. Efforts like these are not futile. Previous research in the creative industries has paid
attention to several prominent award ceremonies and festivals – of which Cannes is an example – and has repeatedly found that being nominated for or winning an award is beneficial for performance (e.g. Anand & Watson, 2004; Mezias, Strandgaard Pedersen, Svejenova & Mazza, 2008; Deuchert, Adjamah & Pauly, 2005; Ginsburgh, 2003; Nelson, Donihue, Waldman & Wheaton, 2001).
In the creative industries, where the quality of a product can only be determined after
consumption, market signals play an important role in exchange decisions (Nelson, 1970).
Because quality cannot be determined in advance, audiences rely on market signals: certain
aspects or behaviors of a product or its producer, that provide information about unobserved
quality (Spence, 1973). In this context, many market signals have been studied extensively,
such as word-of-mouth (e.g. Craig, Green & Versaci, 2013; Kim, Park & Park, 2013),
reputation (e.g. Jensen & Roy, 2008; Rao, 1994; Rindova, Pollock & Hayward, 2006), status
(e.g. Ertug & Castelluci, 2015; Ertug, Yogev, Lee & Hedström, 2016), and certification
contests, of which film festivals are an example (e.g. Anand & Watson, 2004; Deuchert et al.,
2005; Gemser, Leenders & Wijberg, 2008)
Central to this research is the assumption that the effectiveness of market signals will
market signals are Ertug et al. (2016), Gemser et al. (2008), and Kim and Jensen (2013), who
all studied how the relationship between market signals and success was altered by several
moderating variables. The goal of this study is to examine what influences the effectiveness
of participation in a film festival as a market signal in relation to foreign performance of
movies. Market signals that have a positive effect on domestic performance of a movie do not
always have the same influence on performance in foreign markets (Kim & Jensen, 2013).
Kim and Jensen (2013), for example, found that domestic commercial performance and
artistic acclaim had a different effect on export likelihood and foreign commercial
performance, depending on distributor type and the cultural distance between audiences.
They also found that with an increase in cultural distance, critical acclaim – such as movie
awards – became more important, supporting this paper’s claim that film festivals play an
important role in foreign success. In contrast to prior research, this study will look at how
different types of festivals differently influence foreign success.
The distinction between different types of film festivals will be made on two levels.
First, film festivals will be ranked by their status level. Prior research has shown that
certification contests establish a new social hierarchy within a field and that the status of the
organization of these contests will be transferred to its affiliates in the process (Rao, 1994).
This study assumes that participation in a film festival – a certification contest – positively influences foreign performance, because it increases the movie’s status level. Therefore, a film festival with a higher status level is expected to have a different effect on success than a
festival of lower status.
Second, movies as well as film festivals in the sample will be categorized as being
mainstream or art house or somewhere in between. Mainstream movies are widely released
and targeted at a broad audience, whereas art house movies usually serve a niche market and
al. (2008), and Kim and Jensen (2013) all found that certain incongruences can negatively
influence the effectiveness of market signals. For example, a mismatch between a producer’s reputation and an investor’s orientation (Ebbers & Wijnberg, 2012), the incongruence between an artist’s reputation for artistic quality and his or her interaction with commercially oriented galleries (Ertug et al., 2016), a mismatch between the composition of an award jury – either experts, peers, or end consumers – and the audience’s usual source of information, also either experts, peer producers, or themselves as consumers (Gemser et al., 2008), and a
widely distributed, commercially successful movie, that receives an award for artistic quality
(Kim & Jensen, 2013). This research investigates the effect that combining different market
signals has on performance. Specifically, it will examine the effect of a match or mismatch
between movie and festival type – art house or mainstream – on foreign performance.
The goal of this study is, on the one hand, to examine the mediating role of film festival performance in the relationship between a movie’s domestic performance and foreign performance. Domestic artistic performance – in the form of professional critics’ reviews and
online consumer reviews – is expected to influence a movie’s festival performance – for
example, for how many film festivals the movie is selected – because festival organizations
use information about domestic performance to make their selection. Subsequently, festival
performance is expected to influence foreign performance, because it increases a movie’s
status and, therefore, decreases the risk for foreign stakeholders that comes with buying the movie’s rights (Pollock, Chen, Jackson, & Hambrick, 2010). Moreover, because a film festival is a place where different industry stakeholders interact and make deals – for
example, to organize a theatrical release in a foreign country – film festival performance
might influence foreign performance (Mezias et al., 2008). Because thousands of movies are
produced in Europe each year and it is hard for one movie to stand out, film festival
foreign performance. Because online consumer reviews are often in English and easily
accessible for all stakeholders around Europe, this mediation is expected to be stronger in the case of professional critics’ reviews. On the other hand, the goal of this study is to examine whether festival status and congruence between movie and festival type moderate the
relationship between festival performance and foreign performance. In other words: do
higher festival status and a higher level of congruence between movie and festival type
increase the effectiveness of festival performance as a market signal? This study, in the
empirical context of the movie industry, makes several contributions to existing research on
market signal effectiveness in the movie industry.
Firstly, every year large amounts of money are spent on efforts to get a creative
product to participate in a certification contest or to possibly win an award. Little is known
yet, however, about the effect different kinds of certification contests or other so-called field
configuring events have on foreign performance, and how the match between these events
and the creative product influences this relationship. Therefore, this research could tell us
more about whether an investment in these efforts – in this case to be selected for a film
festival – will be useful.
Secondly, a distinction is made between field configuring events on the basis of their
status level. Mezias et al. (2008) studied the influence of European premier film festivals on
overall commercial success of movies and found that a certain classification system exists
among these festivals, with the highest in rank being the most beneficial for commercial
success. However, their research only involved the film festivals of Cannes, Berlin and
Venice, whereas many more festivals take place in Europe every year. This research will
include all film festivals for which movies in the sample were selected, and it will, therefore,
Thirdly, unlike previous research, a distinction is made between film festivals on the
basis of their position on the art house versus mainstream continuum. Instead of studying the
overall mediating effect of film festival performance in the relationship between domestic
performance and foreign performance, this paper will examine whether different festivals
differ in their effect on foreign performance, and in particular if a higher level of congruence
between the movie and the festival will positively influence foreign performance. Previous
research has used the art house versus mainstream division between movies to study the
effect of film reviews on box office performance (Gemser, van Oostrum & Leenders, 2007)
and to study the effectiveness of movie awards based on a match between the kind of award jury and the audience’s usual source of information (Gemser et al., 2008). These studies, however, did not yet apply the art house/mainstream division to film festivals.
So far, research in the creative industry has only focused on a handful of famous field configuring events – such as film festivals and award ceremonies – and has failed to examine how different kinds of events can differently influence performance. It is important to study the effect all these different events have on performance of products in the creative industry, because they cannot be expected to have the same influence as the handful of prominent events that have been studied in the past. Therefore, research is needed to establish how different characteristics of field configuring events – such as their status or congruence with the product – moderate the effect these events have on performance. This research will focus on movies and will include many different European film festivals to investigate how they influence Dutch movies’ foreign performance, based on the festival’s status level and the level of congruence between the movie’s and the festival’s art house versus mainstream level. The sample will consist of Dutch movies released in the Netherlands between 1998
and 2010 and all non-Dutch European film festivals for which they were selected, at which
This paper will proceed as follows. First, background theory will be presented to
support the hypotheses. Second, the data and methods used in this research will be described,
followed by a description of the results. This paper will end with a discussion section and a
2. Theory
The theory section will start with a discussion of market signals that are important in the
creative industries and of the effect of combining different signals. Then, the liability of
foreignness and its effect on success in foreign markets will be introduced. The section will
proceed with a description of field configuring events as conduits of market signals, and of
film festivals as field configuring events. Lastly, the importance of festival status and
congruence will be discussed. The hypotheses will be presented throughout these parts.
2.1 Market signals in the creative industries
Creative products are often experience goods (Nelson, 1970), which means that their quality
can only be evaluated after consumption. This poses a problem for audiences. Because the
quality of a creative product cannot be assessed in advance, audiences instead have to rely on
market signals. Market signals are certain aspects of a product or firm that – unintentionally
or on purpose – convey information about otherwise unobserved quality (Spence, 1974 in:
Kim & Jensen, 2013). These signals play an important role in exchange decisions because they reduce uncertainty about a product’s or firm’s quality. Previous research, conducted in the creative industries, has paid a lot of attention to which market signals exist and how they
work. This research will focus on two of the most widely studied market signals, which are
reputation (e.g. Ertug, Yogev, Lee & Hedström, 2016; Jensen & Roy, 2008; Rao, 1994;
Rindova, Pollock & Hayward, 2006; Tomaselli, Ebbers & Torluccio, 2016), and status (e.g.
Ertug & Castelluci, 2015; Ertug et al. 2016; Jensen & Roy, 2008; and Rindova et al. 2006).
First, reputation is an example of an indirect signal of quality. Jensen and Roy (2008) define reputation as: ‘The prestige accorded actors on the basis of how they have performed particular activities in the past’ (p. 497). Several studies have shown that a good reputation
increases firms’ commercial performance, because it improves their ability to attract resources from various stakeholders – such as investments or purchases – and is therefore a
valuable intangible asset (Petkova, 2016; Rindova et al., 2006). In the movie industry, the
reputation of producers of a movie based on the success of their previous movies can benefit
the success of the current movie – in the domestic market, as well as in foreign markets.
Moreover, because movies are usually released in foreign markets after they have been
released in their domestic market, performance in the domestic market can be viewed as
performance in the past. In the creative industries, two kinds of reputation can help in
attracting valuable resources: (a) a commercial reputation and (b) an artistic reputation. High
commercial performance in the past gives products or producers a commercial reputation.
High commercial performance – high box office revenues in the case of movies – indicates
that many people have already decided to purchase a product – or see a movie – and, thus,
makes your own purchase decision less risky. If the movie was of low quality, it would have
become clear already and box office revenues would not have been so high. An artistic
reputation, on the other hand, results from high critical acclaim. Critical acclaim can mean
several things. It can indicate receiving favorable reviews by either experts or consumers and
it can also mean winning an award. When the acclaim is primarily based on artistic grounds,
it is also called artistic acclaim (Gemser et al., 2007).
Previous research has found that commercial success – high box office – in the
domestic market gives a movie a commercial reputation in the eyes of foreign stakeholders
and will, consequently, have a positive influence on foreign performance (Elberse &
Eliashberg, 2003).
Similarly, previous research has argued that artistic success – high average reviews – in the
domestic market gives a movie an artistic reputation in the eyes of foreign stakeholders. This
market, will positively influence foreign performance of a movie. This leads to the following
hypotheses:
H1a. There is a positive relationship between professional critics’ reviews in the domestic market and the performance of movies in foreign markets.
H1b. There is a positive relationship between online consumer reviews in the domestic market and the performance of movies in foreign markets.
Although this study hypothesizes that professional critics’ reviews and online consumer
reviews will both positively influence foreign performance, these relationships are not expected to be of equal strength. Domestic critics’ reviews are mostly read by the domestic audience, since they appear in – Dutch – newspapers, and are, therefore, only available and
understandable to domestic consumers. Foreign distributors, as experts in the movie industry,
might, however, take an interest in these ratings before releasing a Dutch movie in their
country. Online consumer reviews, on the other hand, are visible on the internet, and are,
therefore, available and understandable to a much larger audience throughout Europe
(Elberse & Eliashberg, 2003). Because they are more easily available, they are expected to have a stronger influence on foreign performance than professional critics’ reviews, which results in the following hypothesis:
H1c. The abovementioned relationship is stronger for online consumer reviews than for professional critics’ reviews.
Second, a market signal that is also indirect and closely related to reputation is status, defined by Jensen and Roy (2008) as: ‘The prestige accorded actors because of the hierarchical
position they occupy in a social structure’ (p. 496). As opposed to reputation, status is not determined by prior behavior alone. Instead, it is based on a comparison between the actor’s prior behavior and the prior behavior of his competitors (Podolny, 1993). The actor that
showed the highest performance in the past, will have the highest status, whereas the actor
with the lowest past performance – compared to his competitors – will have the lowest status. Ertug et al. (2016) describe status as follows: it is ‘based on an actor’s position in a social hierarchy that both reflects and influences how the actor is acknowledged by others’ (p. 11).
This argument resembles that of Podolny (1993), who argues that status is not only an important quality signal for ‘non-producing market participants’ (p. 830), but also determines what opportunities are available to the producer in a market, and consequently how he will
perform. Other stakeholders are more inclined to invest in or collaborate with a high-status
actor than with a low-status actor, because high status signals that this collaboration is less
risky. A high status cannot be obtained individually by producers, but requires a third party to
make a comparison between or ranking among several competitors. So in contrast to
reputation, obtaining status requires affiliation with a higher-status actor. Affiliation with a
higher-status actor enhances a producer’s status, whereas ties to a lower-status actor detract
from it (Faulkner, 1983)
2.2 Combinations of and interactions between market signals
The above described market signals can each individually affect a movie’s performance, but previous research has found that in many cases the impact on performance is the result of an
interplay between multiple signals, as the following section will describe.
Firstly, different signals can impact performance in a particular sequence. For
example, in their research about exchange partner choice in the U.S. audit industry, Jensen
choice process. Specifically, they state that status determines which firms will be admitted to
the active choice set – which will only contain firms from a specific status bracket – whereas
reputation determines the actual partner choice from this set of possible candidates.
Secondly, the effect a market signal has on performance can be based on an
interaction with a third factor. Ebbers and Wijnberg (2012), for example, found that the performance based reputation of a new venture’s founding team member had a stronger effect on the venture’s ability to obtain investment capital when this reputation matched the type of selector the investor was. A new venture was more likely to obtain investment capital from
investors that were market selectors when a founding team member had a market reputation
(Ebbers & Wijnberg, 2012). Interestingly, they found that this match was also important in a
negative sense, so that obtaining investment capital became less likely when there was no match between the founding member’s reputation and the type of selector. Similarly, a match between different signals can moderate the impact a particular market signal has on
performance. Ertug et al. (2016), in their research in the contemporary art field, examined how the relationship between an artist’s audience-specific reputation and success was moderated by the level of contingency between this reputation and the artist’s status. For example, when an artist has a reputation for artistic quality and his work is exhibited in a
museum – which is believed to be of higher status than a commercially oriented gallery – the
level of congruence is high. When the same artist, however, exhibits his work in the
commercially oriented gallery – which is of lower status than the museum – the congruence
level is low. The level of congruence will also be high when an artist with a commercial
reputation exhibits in a gallery. Ertug et al. (2016) also found that such a mismatch between
status and reputation – for example, an artist with an artistic reputation exhibiting in a gallery – could have a negative influence on success (Ertug et al., 2016).
Another example where a mismatch between signals proved to have a negative effect
on performance is the research by Gemser et al. (2008). They focused on the U.S. motion
picture industry and the effect of movie awards on box office revenue and share of screens,
and found that this relationship was moderated by the composition of the award jury. These
juries could be composed of either experts, peers, or end consumers – for example, movie
experts, other industry participants, or moviegoers. They found that for independent movies
the award had a more positive effect on box office revenue and share of screens when the
jury was composed of experts, rather than peers or end consumers. The match between the
jury composition and the independent movie audience’s usual source of information – also
experts – made the award a more effective market signal than an award selected by end
consumers. For mainstream movies, they did not find the same effect.
Important stakeholders in the movie industry are producers, distributors, exhibitors and
audiences. Movies are produced by either – large – production studios or by independent
producers. Movie distributors buy movie rights and sell them to exhibitors, who can then
screen the movie in their theatre where audiences can see it.
A distinction often made in the movie industry is the one between independent and
major, or art house and mainstream. Art house movies, for example, focus on a niche market
and are distributed on a small scale, whereas mainstream movies aim to reach a broad
audience and are widely distributed. The distinction often also indicates a difference between,
on the one hand, artistic quality (art house) and, on the other hand, commercial viability
(mainstream) (Gemser et al., 2007) and besides movies, it can also be applied to distributors,
2.3 Success in foreign markets: The liability of foreignness
Market signals are important for the performance of movies in their domestic market,
because their quality can only be determined after consumption. When movies are distributed
in foreign markets, however, they face an audience of foreign distributors, exhibitors and
moviegoers and, as a consequence, a new set of problems. These problems are caused by the
fact that for creative products market signals that are effective in the domestic market do not
automatically translate to and, thus, have the same effect on foreign audiences (Kim &
Jensen, 2013; Lee, 2009). Unlike refrigerators and cellphones, creative products are often deeply embedded in a country’s culture. Understanding and appreciation of these products, therefore, depends on understanding and appreciation of this particular culture. According to
Lee (2009), this is the reason why people prefer creative products that are embedded in a
culture they are familiar with. When a creative product that is strongly influenced by culture enters a foreign market, its value is, therefore, ‘discounted’ (Lee, 2009). Kim and Jensen (2013) called this effect the liability of foreignness and found that it gets stronger when the ‘distance’ between two cultures gets larger, so for a Dutch movie it is even more difficult to appeal to a Japanese audience than it is to an audience in Belgium, because Japan is culturally
more distant from the Netherlands than Belgium.
Because people’s taste in movies is partially determined by what is easy for them to understand, signals of quality that were meaningful for the domestic audience, will not
necessarily be as meaningful to a foreign audience. For example, high commercial
performance in the Dutch market means that many Dutch consumers liked the movie, but this
does not automatically mean that a foreign audience will appreciate it as well (Kim & Jensen,
2013). So what market signals do foreign audiences use to acquire information about a creative product’s quality?
2.4 Field configuring events as conduits of market signals
In the context of institutional theory, DiMaggio and Powell define a field as “those
organizations that, in the aggregate, constitute a recognized area of institutional life: key
suppliers, resource and product consumers, regulatory agencies, and other organizations that produce similar services or products” (DiMaggio & Powell, 1983: p. 148). Fields, like the movie industry, are continuously changing. For example, they can be shaped or formed by
field configuring events. Meyer, Gaba and Colwell (2005) define field configuring events as “settings where people from diverse social organizations assemble temporarily, with the conscious, collective intent to construct an organizational field” (p. 467). Field configuring events (FCEs) are temporary gatherings of actors from a particular field, which offer the
opportunity for formal and informal social interaction and exchange of information. They
offer a setting in which new ideas can be shared and accomplishments can be recognized.
They accommodate the formation and consolidation of network ties and are, therefore, a way
to generate valuable resources (Lampel & Meyer, 2008). Examples of FCEs are industry
events, certification contests, and award ceremonies.
Prior research has found that participating in FCEs can have a positive impact on
performance (e.g. Lampel et al., 2008; Meyer et al., 2005; and Stam, 2010). This study argues
that FCEs have a positive influence on performance because they are conduits of market
signals, and in particular of status. As mentioned before, status cannot be obtained
individually, but requires affiliation with other – higher-status – actors. Podolny (1993)
argues that the relationship between quality and status is mediated by social relations. FCEs
are generally organized by special-purpose organizations. Affiliation with these organizations – which could, for example, mean being invited to an industry event – changes other stakeholders’ perceptions of the actor. If the organization is of high status, this means the
perceived status of the actor will also be higher (Podolny, 1993). This ‘transfer’ of status works in two directions: when a high-status actor affiliates with a low-status actor, its status
will be lower. Organizations of FCEs will therefore choose carefully with whom they affiliate. When it is hard to assess a product’s quality before consumption, as is the case for experience goods, stakeholders will use this kind of judgment as an indication of quality. In
this way, status allocation becomes a self-reinforcing process: as soon as an actor is perceived
and noticed by another actor as being of high status, other stakeholders, who are unable to assess the actor’s underlying quality, will be influenced by this judgment. And so status – perhaps originally based on underlying quality – will be amplified because of a social process
(Gould, 2002).
Film festivals are events that usually take place in one city or even in one theatre and
which differ in length from one day to a couple of weeks. A festival has a jury, that can be
composed of industry experts, consumers, or, for example, industry artists, such as directors,
screenwriters, or actors. This jury selects a number of movies, that they found remarkable
over the last year, to be screened during the festival and offer a local or an international
audience the opportunity to see these movies. It also happens that movies have their world
premiere at a film festival. Many film festivals also have an award ceremony. Several months
to weeks before the festival takes place, nominations for the awards are made public and
during a special ceremony – often at the end of the festival – the winners are announced and
receive their award (Rüling & Strandgaard Pedersen, 2010). Film festivals are a place where
various industry stakeholders come together and interact. Not only movie consumers and
nominated artists and producers are present, but also movie critics, distributors, exhibitors,
and peer artists and producers will use the festival as an opportunity to witness recent
developments in the fields, gather new contacts, and arrange business deals (Mezias et al.,
and are probably the most important one in the movie industry. They are taking place all over
the world in many different shapes and sizes (Rüling & Strandgaard Pedersen, 2010). Europe’s most famous film festivals are the Venice International Film Festival, the Cannes Film Festival, and the Berlin Film festival, however, many more exist.
Performance of a movie at a film festival can mean three things: (a) a movie can be
selected for a film festival, (b) a movie can be nominated for an award at a film festival, and
(c) a movie can win an award at a film festival. The organization of the festival and – in the
case of an award ceremony – the award jury will decide which movie is selected, which
movie is nominated for an award, and which movie wins the award. Presuming they will see the movies first, their judgment will of course be based on their perceptions of the movies’ quality. Each year, thousands of movies are produced in Europe and festival juries will not see them all. To be featured on a festival’s long-list it is, therefore – although difficult – very important for a movie to stand out. One way in which a movie can catch a festival’s attention is by receiving high average reviews in its domestic market. This study therefore argues that a movie’s performance in the domestic market – in the form of reviews – positively influences its film festival performance.
Recall, however, the study by Kim & Jensen (2013), who stated that appeal of a
movie in its domestic market is not always transferable to a foreign market, because of
cultural differences, and that commercial success in the domestic market is, therefore, not
always a good predictor of commercial success in a foreign market. They also stated that
artistic acclaim - or professional critics’ reviews – was a better predictor of success in a
foreign market, because this is decided upon by a group of movie experts, who are more
likely to share a universal set of conventions and criteria for assessing the quality of a movie,
because they have experience in the movie industry and often have a formal education in film
better indicator than high commercial performance of whether or not a movie’s appeal is transferable to a different market. This research hypothesizes that a movie’s performance in its domestic market positively influences film festival performance, but that this effect is
stronger for movies with high artistic performance than for movies with high commercial
performance in the domestic market.
H2a. A movie’s domestic commercial performance positively influences a movie’s film festival performance.
H2b. A movie’s domestic artistic performance positively influences a movie’s film festival performance.
H2c. For art house movies, the abovementioned relationship is stronger for domestic artistic performance than for domestic commercial performance.
H2d. For mainstream movies, the abovementioned relationship is stronger for domestic commercial performance than for domestic artistic performance.
2.5 Film festivals as conduits of market signals
This study proposes that film festival performance increases a movie’s status and, therefore,
that film festivals are conduits of market signals. Film festivals are field configuring events
that help establish a new hierarchy in the field of movies, by selecting movies of high quality
and awarding the ones of excellent quality, while low quality movies are excluded. Pollock et
al. (2010), in their research on young firms with prestigious affiliates, provide two reasons
why other industry stakeholders use this hierarchy as a basis to their investment decisions.
First, these high-status actors – film festival organizations and award ceremony juries – are expected to have ‘superior abilities’ to determine with whom they should affiliate (Pollock et al., 2010: p.9). Just like Kim and Jensen (2013) described, their own expertise makes their
judgment more trustworthy. Second, the transfer of status takes place in two directions and so
a high-status actor can suffer from affiliation with a low status actor (Podolny, 1993), or as Pollock et al. (2010) put it, ‘prestigious actors value their status highly’ (p. 9) and will not put it at stake by making arbitrary decisions. Therefore, an organization’s decision to select a movie for their festival signals that this particular movie is worthy of their affiliation and fits
within their status bracket.
Because participation in a film festival increases a movie’s status – as this study assumes – foreign stakeholders will be more likely to invest in it, because status decreases the
uncertainty of an investment decision. This study, thus, hypothesizes that participation in a
film festival will have a positive influence on foreign performance.
H3. There is a positive relationship between a movie’s film festival performance and its foreign performance
Film festivals are conduits of market signals. To a foreign audience they provide information about a movie’s status and, in part, about its underlying quality. Therefore, this study hypothesizes that film festival performance at least partially mediates the relationship
between a movie’s performance in the domestic market and its foreign performance.Because
online consumer reviews are easily accessible for everyone in Europe, they are expected to
have a direct influence on foreign performance – as stated in hypothesis 1b. Therefore the
mediating effect of festival performance in the relationship between domestic performance and foreign performance is expected to be stronger for professional critics’ reviews than for online consumer reviews.
H4a. The relationship between professional critics’ reviews in the domestic market and the performance of movies in foreign markets is mediated by a movie’s film festival performance. H4b. The relationship between online consumer reviews in the domestic market and the performance of movies in foreign markets is mediated by a movie’s film festival performance. H4c. The abovementioned relationship is stronger for professional critics’ reviews than for online consumer reviews.
2.6 The importance of film festival status
This research assumes that film festivals are conduits of market signals, and particularly of status, and that film festival performance increases a movie’s status. Because it increases a movie’s status, film festival performance should have a positive influence on foreign performance. As we saw earlier, status can be obtained by affiliating with a higher-status actor. This affiliate’s high status is, then, as it were, transferred to the focal actor (Podolny, 1993). Thus, it can be expected that the higher the affiliate’s status, the higher the actor’s
status will be. In other words, when a movie is selected for a high-status film festival, its
perceived status will be higher than when it is selected for a lower-status film festival. And the higher a movie’s status, the higher its foreign performance. In sum, this research hypothesizes that film festival status positively influences the relationship between film
festival performance and foreign performance.
Moreover, this effect is expected to be stronger for art house movies than for
mainstream movies. As described earlier, commercial success of a movie in one country is
not automatically transferable to a foreign market, whereas artistic acclaim – coming from
film experts that share a set of more universal values and knowledge – is generally more
valuable to a foreign audience (Kim & Jensen 2013). Moreover, the audience for art house
word-of-mouth is small, compared to mainstream movies (Gemser et al., 2007). It is expected,
therefore, that status – affiliation of and recognition by experts – is valued higher by art house
audiences than by mainstream audiences.
H5. Festival status positively influences the relationship between a movie’s film festival performance and foreign performance and this effect is stronger for art house than for mainstream movies.
2.7 The importance of congruence
The research by Ertug et al. (2016), Gemser et al. (2008), and Kim and Jensen (2013) showed
that under certain conditions a market signal can be more or less effective. They found, for
example, that incongruence of signals can negatively influence success. Kim and Jensen
(2013) argued that for foreign mainstream movies, that are distributed by a major distributor,
artistic acclaim can have a negative effect on success. A movie with major distribution is
initially expected to appeal to a broad audience, but when it receives an award – a signal of
artistic quality – audiences might suddenly expect it to be too ‘artsy’ and will be reluctant to
see it (Kim & Jensen, 2013). Moreover, Ertug et al. (2016) showed that when an artist with a
reputation for artistic quality interacts with – commercially oriented – galleries, his or her
chances to exhibit in a high status museum diminish. When different market signals are mixed, it seems their effect on a creative product’s performance decreases. Ertug et al. (2016) state that congruence between status and reputation is more important for an audience that
faces artistic accountability than for an audience that faces commercial accountability.
Similarly, Gemser et al. (2008) found that congruence between an award jury and the audience’s usual source of information was more important for art house audiences than for mainstream audiences. Their study showed that art house movies benefited more from
winning an expert-selected award than from winning a peer-selected or market-selected
award, whereas for mainstream movies, the effects of winning these different kinds of awards
did not differ significantly (Gemser et al., 2008).
The distinction between art house and mainstream producers, distributors, exhibitors,
and audiences can also be applied to film festivals. Art house film festivals would then, for
example, be festivals that mostly screen art house movies or have an art house character and
mainstream film festivals would mostly screen mainstream movies and have a mainstream
character. Kim and Jensen (2013) state that having a reputation for commercial viability and subsequently winning an artistic award has a negative effect on audience’s willingness to see a movie, because these mixed messages create uncertainty about what can be expected of a
movie. However, not all awards and film festivals are artistically oriented. There are also ‘mainstream’ awards and festivals, where, for example, consumers decide which movie wins an award. This means that a movie could also be artistically successful and subsequently win
a mainstream award. Because previous research has shown that the influence of market
signals on performance diminishes when different signals are mixed, this study hypothesizes that congruence between a movie’s and a film festival’s art house or mainstream level will positively influence foreign performance. And because for an art house audience congruence
seems to be more important than for a mainstream audience (Ertug et al., 2016; Gemser et al.,
2007) this effect will be stronger for art house movies than for mainstream movies.
H6. Congruence between a movie’s art house/mainstream level and a festival’s art house/mainstream level positively influences the relationship between a movie’s film festival performance and foreign performance and this relationship is stronger for movies with a high art house level.
Figure 1. Conceptual framework Domestic performance Film festival performance Foreign performance Congruence Status
3. Data and method 3.1 Research setting and data collection
The empirical setting of this research is the European movie industry. The focus of this
research is the effect of film festival performance - being selected, being nominated for an
award or winning an award – of movies on their performance outside the domestic market
and within Europe – indicated by foreign releases and foreign admissions. The sample
consists of Dutch movies that were released in The Netherlands between 1998 and 2010. A movie is considered ‘Dutch’ when its producers and directors have a Dutch nationality or have attended the Dutch Film Academy (Ebbers & Wijnberg, 2012). A search in the Lumiere
database of the European Audiovisual Observatory (http://lumiere.obs.coe.int/web/search/)
yielded a result of 342 movies. 55 movies were excluded, because they were a coproduction
of different countries. When a movie is produced by companies from more than one country,
it is more likely that it will be released in both countries. Moreover, chances are that it will
contain traces of all these countries’ cultures and that it will appeal to a larger audience than
just the Dutch moviegoers (Kim & Jensen, 2013). The fact that a movie is a coproduction is,
therefore expected to positively influence performance. After excluding coproductions,
documentaries, and short films, the final sample consisted of 191.
3.2 Dependent variables
The dependent variable of this study is the performance of Dutch movies in non-Dutch
European countries. This foreign performance is measured in two different ways. Firstly,
foreign performance is measured as the total number of non-Dutch European countries in
which the movie is released, data which is also available in the Lumiere database. As an
exception, a release in Belgium was not considered as a foreign release, because Belgium and
produced for the Dutch and the Belgian market at once. Secondly, the total foreign
admissions in all non-Dutch European countries in which the movie is released, is used to give an impression of the movie’s overall performance in Europe.
3.3 Independent variables
Performance in the domestic market – The performance of a movie in its domestic
market is measured in three different ways. Firstly, a movie’s commercial performance in the
domestic market is measured as its total admissions. Data for this measure was collected
using the Film Fund database, which provides information about admissions. The Film Fund
is a cultural fund, installed by the Dutch government, to stimulate the production of diverse,
high-quality movies and to provide an independent production climate in the Netherlands for
the art of film. Secondly, a movie’s artistic performance in the domestic market is measured as (a) the average review the movie received from movie critics’, and (b) the average review the movie received from online consumers. Critics’ ratings were retrieved from the Filmkrant, a Dutch magazine about film, which presents a movie’s average rating, based on
all reviews that appeared in premium newspapers. The rating a movie has received on IMDb
is used as the average online consumer rating. IMDb is an online database of movies,
television programs and video games, with information about cast, production crew,
characters, plots, and reviews.
Film festival performance – Film festival performance is operationalized as the
number of times a movie was selected for a non-Dutch European film festival, the number of
times a movie was nominated for an award at a non-Dutch European film festival in the ‘best film’, ‘best director’, or ‘best script’ category, and the number of times a movie won an award at a non-Dutch European film festival in one of these three categories. Information
Mean film festival status – A status scale was used, ranging from 1-6. All film
festivals were ranked on this scale by two industry experts, who were familiar with the world
of international film festivals. Both experts were working at a Dutch film production
company. Their rankings were compared. To calculate the mean status of the festivals for
which the movie was selected, all values for status together were divided by the number of
non-Dutch European festivals for which the movie was selected.
Mean congruence between movie’s and festival’s mainstream/art house level – The
congruence between a movie’s and a festival’s mainstream/art house level is the absolute difference between these two levels. A low value for congruence, therefore, indicates a
higher level of congruence, because the absolute difference is small. To calculate the mean congruence between the movie’s and the festival’s art house/mainstream level, all absolute values for congruence together were divided by the number of non-Dutch European festivals
for which the movie was selected. The minstream/art house levels for the movies and
festivals was derived as follows.
All movies in the sample were rated on a scale from 1-6 to indicate their
mainstream/art house level. This rating was based on the number of opening screens a movie
had in the domestic market. Mainstream movies, which are expected to appeal to a large
audience, are usually released in many places at the same time, and therefore have a larger
number of opening screens than art house movies, which are aimed at a much smaller niche
audience. Information about the number of opening screens was available in the Film Fund
database. After the information about the number of opening screens was collected for all
movies in the sample, they were divided into six equal groups based on this number.
All festivals for which movies in the sample were selected, at which they were
nominated for an award or won an award are also rated on a scale from 1-6, ranging from
The mean of their ratings was used to make the final rating. When only one expert rated the
festival, this rating was used. Both raters were experts in a slightly different field, namely the field of children’s movies and festivals and the field of fiction movies and festivals. Therefore, their ratings often differed. The expert of fiction movies, for example, allocated a low status score to some festivals that the children’s movies expert declared to be the most important and influential festivals in the children’s movies industry, and vice versa. In these cases, I decided to still use the mean of the two ratings for the final rating. This final rating, however, still contained many missing values, because many festivals were unknown to the raters. For these missing values, an estimation was made. In the interview, one of the experts stated that there are only a few important film festivals, that are widely known. Apart from these few high-status festivals, there are many other unknown film festivals, that are often aimed at a very specific niche market. Being aimed at a niche market is one of the characteristics of art house movies, and, therefore, the festivals that were unknown to the two experts were rated on the art house/mainstream scale with a 6, meaning their art house level was very high. Because the experts considered being unknown as being of low status, the festivals with missing values in the sample were rated with a 1 on the status scale.
3.4 Control variables
Several control variables were included in the analysis to represent different forces in the
movie industry that could have influenced the model.
First, I included several dummy variables for genre, respectively comedy, action, family,
drama, thriller, romance, and crime. The category other genre was used as the baseline
category. Previous research has found that genre plays an important role in audiences’ movie choice, because it provides them with information about the movie’s characteristics, such as its storyline and characters (Hsu, 2006).
Second, although the Oscars are an award ceremony and not a film festival, an Oscar
nomination for best foreign movie and an Oscar win for best foreign movie are included as
control variables, because previous research has demonstrated that being nominated for or winning an Oscar can significantly boost a movie’s performance (e.g. Deuchert et al., 2005; Nelson et al., 2001).
Third, a dummy variable was used to indicate not English spoken movies. Language is an
important component of culture and the chance that movie audiences in Europe understand
English is significantly larger than that they understand Dutch. English as the main language in a movie can, therefore, decrease the liability of foreignness and increase a movie’s performance abroad (Kim & Jensen, 2013).
Finally, original script was included as a dummy variable, because scripts that are
non-original and are, for example, based on a book, are expected to positively influence a movie’s
performance. Because the quality of the book is known in advance, the purchase decision of
4. Results
In the following section the results are presented. The result section is structured as follows:
first, descriptive statistics of the sample and a full correlation matrix – containing all
variables – will be presented; second, to test hypotheses 1 through 4, the results of the
mediation analysis will be discussed; third – to test hypotheses 5 and 6 – the results of the
moderated mediation analysis will be discussed; and fourth, several robustness checks that
were performed will be described.
4.1 Descriptive statistics
Descriptive statistics and correlations for all variables are provided in table 1. Normality
checks were performed to see if the variables had a normal distribution. The variables
domestic market review, number of non-Dutch European festivals for which selected, number of non-Dutch European countries in which released, total foreign admissions and the control
variable number of non-European festivals for which selected all had a positively skewed and
leptokurtic distribution. Before using these in the analyses I performed a log transformation
on these variables. Missing values were dealt with by excluding them pairwise.
Art house/mainstream levels and status levels for all film festivals in the sample were
rated by two experts. Reliability was measured using only the festivals that were rated by
both the experts and was found to be rather low. The art house/mainstream level ratings had a
Cohen’s Kappa of = 0.003, with p > 0.05. The Cronbach’s alpha of the scale was also low ( = .38). The status level ratings had a Cohen’s Kappa of = 0.302, with p < 0.01, and a
Cronbach’s alpha of .776. This is a lot higher than the art house/mainstream ratings, but Cohen’s Kappa is still quite low. Possible reasons for the low interrater reliability score for both measures will be presented in the discussion.
Table 1: means, standard deviations, correlations
Variables M SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
1. Not English spoken .98 .14 -
2. Foreign distributor .15 .36 -.04 - 3. Original script .72 .45 -.01 -.06 - 4. Oscar nomination .01 .07 .01 -.03 .05 - 5. Comedy genre .29 .46 .09 .05 .15* .11 - 6. Action genre .05 .21 .03 -.10 .03 -.02 .08 - 7. Family genre .20 .40 -.02 .05 -.25*** -.04 -.17** -.05 - 8. Thriller genre .09 .29 .05 -.03 -.13 -.02 -.20** .02 -.16* - 9. Romance genre .14 .35 .06 .00 -.09 .18* .29*** -.09 -.16* -.02 - 10. Crime genre .08 .27 .04 -.07 .01 -.02 -.02 .30*** -.15* .32*** -.00 - 11. Drama genre .57 .50 -.05 -.07 .08 -.08 -.31*** -.21** -.38*** -.10 -.02 -.06 - 12. # of non-European
festivals for which selected .89 1.88 -.05 .09 .04 .08 .13 -.03 -.12 -.04 .01 .07 16* -
13. Domestic admissions
(x1000) 118.27 192.15 -.05 .04 -.32*** -.03 .16* .11 .21** -.07 .20** -.09 .23*** - .14 -
14. Domestic critics’ review 2.54 .83 .04 -.06 -.10 .02 -.10 -.18 -.01 .00 -.11 .06 .15 .30*** .14 -
15. Domestic market review 6.11 .96 .00 .00 .00 .02 -.15* -.19* -.15* -.06 -.08 -.01 .30*** .22** -.08 .51*** -
16. Total foreign admissions
(x1000) 8.95 48.09 -.06 .03 -.16* -02 .09 .19** .03 -.04 .02 -.04 -.03 .08 .52** .08 .06 -
17. # of non-Dutch European
countries in which released .31 .95 -.08 .07 -.16* .02 .08 -.02 .05 -.10 .14 -.11 .05 .52** .42** .26** .09
.
33** -
18. # of non-Dutch European
festivals for which selected 1.17 1.91 -.06 .14 .09 .18* .05 -.03 -.11 -.09 .07 -.02 .12 .59*** .05 .40*** .30*** .04 .39***
19. Mean Status 2.70 1.65 -.03 .06 -.06 .00 -.07 -.13 .14 -.20 -.05 -.10 .09 .02 -.10 .02 -.07 .05 .12 -.05
4.2 Hypothesis testing Mediation
I performed a regression analysis to test the mediation model, including domestic critics’
review as the independent variable, the number of non-Dutch European countries in which a movie was released as the dependent variable, and the number of festivals for which a movie was selected as the mediator. The following control variables were also included in the
model: not English spoken, foreign distributor, original script, Oscar nomination, number of
non-European festivals for which selected, and the genre dummies comedy, action, family, thriller, romance, crime and drama. None of the movies in the sample won an Oscar, so the
control variable Oscar win, was not included. I tested the same model with domestic market
review as the independent variable, keeping all the other variables the same. In the first
model, I included the variable domestic consumer reviews as a control variable and in the
second model I did the same with domestic critics’ reviews. Figure 2 shows the model that
was used. Results of these analyses are shown in table 2. Note that table 2 contains the results
of three different analyses namely one with domestic critics’ reviews as the independent
variable and the other with domestic consumer reviews as the independent variable. It
therefore contains two different direct effects (c1’). The direct effect of domestic admissions
on foreign performance is not reported in table 2, because only the effect of domestic
admissions on film festival performance (a1) was part of this study.
Figure 2. PROCESS model 4: mediation
X
a1 b1
c1’
Y M
Table 2a. Mediation analysis
Number of non-Dutch European festivals for which selected (M)
Number of non-Dutch European countries in
which released (Y)
Coeff. SE p Coeff. SE p Not English spoken -.02 .12 .88 -.02 .09 .79 Foreign distributor .05 .04 .25 .03 .03 .36 Original script .05 .04 .18 -.04 .03 .21 Oscar nomination .26 .21 .22 .03 .16 .84 Comedy genre -.04 .04 .30 .01 .03 .79 Action genre .09 .08 .25 -.02 .06 .81 Family genre .00 .05 .94 -.02 .04 .65 Thriller genre -.07 .06 .26 -.06 .05 .21 Romance genre .07 .05 .14 .04 .04 .26 Crime genre -.04 .07 .57 -.05 .05 .26 Drama genre -.04 .04 .28 -.01 .03 .64 # of non-European festivals for which
selected .60 .07 .00*** .30 .06 .00*** Domestic critics’ review (X) a1 .09 .02 .00*** c1’ .01 .02 .79 Domestic market review (X) a1 .30 .24 .21 c1’ .09 .18 .62 Domestic admissions (X) a1 -.02 .02 .35 # of non-Dutch European festivals for which selected
(M) b1 .03 .06 .58
constant i1 -.24 .24 .32 i2 -.12 .18 .49
R2=0.54 R2=0.35
F(15,162) = 12.55, p=.00** F(16,161) = 5.50, p=.00**
***P-value is significant at the 0.001 level (2-tailed) **P-value is significant at the 0.01 level (2-tailed) *P-value is significant at the 0.05 level (2-tailed)
Table 2b. Mediation analysis model 1: Direct, total, and indirect effect of domestic critics’ reviews
Effect SE p LLCI ULCI
Direct effect c1' .01 .02 .79 -.03 .04
Total effect c1 .01 .02 .66 -.03 .04
Boot SE
Boot
LLCI Boot ULCI
Indirect effect a1b1 .00 .01 -.01 .02
Table 2c. Mediation analysis model 2: Direct, total, and indirect effect of domestic market reviews
Effect SE p LLCI ULCI
Direct effect c1' .09 .18 .61 -.26 .44
Total effect c1 .10 .18 .58 -.25 .45
Boot SE LLCI Boot Boot ULCI
Indirect effect a1b1 .01 .02 -.02 .10
***P-value is significant at the 0.001 level (2-tailed) **P-value is significant at the 0.01 level (2-tailed) *P-value is significant at the 0.05 level (2-tailed)
The results in table 2b show that there is a positive relationship between professional critics’ reviews in the domestic market and the performance of movies in foreign market – measured
as the number of European countries in which the movie was released – of =.01. This
relationship, however is not significant (with p>0.05 and a 95% BC confidence interval that
is not entirely above zero). Thus, hypothesis 1a is not supported.
The results in table 2c show that the relationship between online consumer reviews in
the domestic market and the performance of movies in foreign markets is also not significant
(=.09, with p>0.05 and a 95% BC confidence interval is not entirely above zero).
Therefore, hypothesis 1b is also not supported. Hypothesis 1c stated that the relationship
between online consumer reviews in the domestic market and performance in foreign markets was stronger than the relationship between professional critics’ reviews in the domestic market and performance in foreign markets. Results show that this is, in fact, the other way
around, but because both effects are not significant, hypothesis 1c is also rejected.
Hypothesis 2 stated that performance of a movie in its domestic market has a positive influence on a movie’s film festival performance. Results shows that the relationship between a movie’s domestic commercial performance and its film festival performance is not significant (=-0.02, p>0.05). Therefore, hypothesis 2a is rejected. On the contrary, the
results do provide support for hypothesis 2b, which states that there is a positive relationship between a movie’s domestic critics’ reviews and its film festival performance. Table 2a shows that this relationship is significantly positive (=0.09, p<0.001).
Hypotheses 2c and 2d state that for art house movies domestic artistic performance
has a stronger relationship on festival performance than domestic commercial performance
and that for mainstream movies it is the other way around. To test these hypotheses the
sample was split in two groups of the same size, based on the movies’ art house/mainstream
level. Subsequently, the same mediation analysis described above was performed on these
separate samples. Results of this analysis can be found in table 3 and 4. Support was found
for hypothesis 2c. For art house movies, the relationship between domestic critics’ reviews
and film festival performance (=0.12**, p<0.001) was stronger than the relationship
between domestic commercial performance and festival performance (=-0.02, p>0.05). This
suggests that festival success for art house movies is influenced more by a positive critics’
rating than by commercial success. For mainstream movies, however, the relationship
between domestic critics’ reviews and film festival performance (=0.06, p>0.05) was the same as the relationship between domestic commercial performance and festival performance
(=0.06, p>0.05). Thus, hypothesis 2d was rejected.
Hypothesis 3 stated that there is a positive relationship between a movie’s film festival performance and its foreign performance. Table 2 shows that the relationship between the
number of non-Dutch European festivals for which a movie was selected and the number of
European countries in which it was released is not significant (=0.03, p>0.05). Therefore,
Hypothesis 4 proposed that festival performance acts as a mediator between performance in
the domestic market and foreign performance. To examine whether this mediation effect is
taking place, several steps were taken. First, table 1 shows that there is a significant
correlation (r=0.26, p<0.01) between the independent variable (domestic critics’ reviews) and
the depencent variable (the number of non-Dutch European countries in which released).
Second, table 1 also shows that there is a significant correlation (r=0.40, p<0.001) between
the independent variable and the mediator (the number of festivals for which a movie is
selected). Third, however, table 2a shows that the mediator does not influence the outcome
variable (b1=0.03, p>0.05). Therefore, it can be concluded that the mediation effect does not
take place.
In the case of domestic online consumer reviews, there is a significant correlation
(r=0.30, p<0.001) between the independent variable and the mediator (the number of festivals
for which a movie is selected), but not between the independent and the dependent variable
(the number of non-Dutch European countries in which released) with r=.09 and p>0.05.
Thus, also in this case mediation is not taking place.
Not surprisingly, table 2b shows that the indirect effect of domestic critics’ reviews
via festival performance on foreign performance is not significant with =0.00 and a 95% BC
bootstrap confidence interval that includes zero (-0.01, 0.02). The same is true for the
indirect effect of domestic consumer reviews via festival performance on foreign
performance, with =0.01 and the 95% BC bootstrap confidence interval that includes zero
(-.02, 0.10). The indirect effect of domestic critics’ reviews is, however, stronger than the indirect effect of domestic market reviews on foreign performance – as stated in hypothesis