The Power of the Press

(1)

MSc Business Administration:

Entrepreneurship & Management in the Creative Industries

The Power of the Press

An empirical study of the influence of newspaper reviews on the success of Broadway musicals with regards of a newspaper outlet’s reputation.

by Rebekka Krempler

10860983 EC 20220620030617

Supervisor: Frederik B. I. Situmeang

Amsterdam Business School, June 24^th, 2022

(2)

Tables and Figures

Figure 1: Conceptual Model ... 29

Figure 2: Newspaper Ranking (as per survey) ... 37

Figure 3: Conceptual Diagram of Model 11 ... 47

Figure 4: Conceptual Diagram of Model 12 ... 47

Figure 5: Scatterplot with Regression Line (All Data) ... 49

Figure 6a-f: Scatterplots with Regression Lines ... 52

Table 1: Number of Reviews per Newspaper Outlet ... 34

Table 2: Variable Summary ... 42

Table 3: Descriptive Statistics of Variables ... 45

Table 4: Correlations Analysis of Variables ... 45

Table 5: Model 1 Summaries ... 48

Table 6: Model 2 Summaries ... 49

Table 7: Effects of SENT on the Number of PERF (Model 1b) ... 50

Table 8: Effects of SENT on the GROSS (Model 2b) ... 51

Table 9: Indirect Effect (SENT à PERF à GROSS) ... 54

Table 10: Indirect Effect (SENT à PERF à GROSS) at the Values of CONT ... 56

Table 11: Model Summaries, Model 12 ... 57

Table 12: Model for PERF (Outcome Variable) ... 57

Table 13: Model for GROSS (Outcome Variable) ... 57

Table 14: Test of conditional X*W interactions at values of Z ... 58

(5)

Statement of Originality

This document is written by student Rebekka Krempler, 10860983, who declares to take full responsibility for the contents of this document.

I declare that this text and the work presented in this document are 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 the completion of the work and its contents.

(6)

Abstract

Many academics have researched the impact of expert evaluations, for instance, newspaper reviews, on the performance of entertainment goods. In their studies, some have even discovered dissimilarities in the influence of newspaper reviews depending on the newspaper outlet. However, scholars have neglected to investigate the underlying reasons for these discrepancies.

Based on the past literature and the fundamentals of Signaling Theory, this paper proposes a moderating effect of a newspaper outlet's reputation on the relationship between a newspaper review on the performance of a theatrical production. Using the example of the Broadway industry, this thesis's data includes the 624 reviews of 125 Broadway musicals with their official opening day between 2007 and 2019. The analysis findings support a moderate positive moderation effect of reputation on the influence of expert review on the success of a musical. Additionally, the results revealed that the supported moderation effect diminishes depending on whether the show is a new production or a revival; the moderation effect becomes negligible for the latter. Lastly, the proposed hypothesis that a similar phenomenon occurs for musicals based on content previously known to the audience was rejected.

This paper provides valuable contributions to the field as it opens the door for further research on the potential factors influencing the effect of expert evaluations in the entertainment industry.

(7)

1. Introduction

“Wicked does not, alas, speak hopefully for the future of the Broadway musical.”

(Brantley, 2003)

“It's safe to say that this is the most complete, and completely satisfying, new musical I've come across in a long time.” (Gardner, 2003)

These two quotes, respectively, stem from The New York Times and the USA Today review of the Broadway musical Wicked, which opened in 2003. Upon reading Brantley's quote, one might conclude that Wicked suffered to stay open on Broadway and has not been very successful since.

However, the exact opposite conclusion might be drawn from Gardner's quote. While the musical received an overall mix of reviews, the two quotes indicate a review's two extreme sentiments:

very positive or very negative. Along with many others, Ben Brantley most likely did not expect the show to become one of the most famous musicals around the globe. With the Broadway production still running to this day, it was translated into multiple languages and was shown to audiences in numerous countries, including the UK, Germany, Australia, and Japan.

Many scholars have focused on different aspects of the impact critics have on the performance of entertainment goods. Even though Brantley’s quote, mentioned above, shows that sometimes critics are unable to foresee success, almost all scholars agree on the apparent correlation between an expert's review and the revenue made (e.g., Hirsch, 1982; Shrum, 1991;

Hirschman and Pieros, 1995). Eliashberg and Shugan (1997) made a tremendous addition to this research field by differentiating between the influencer and the predictor perspective of a critic.

While the first perspective supports a critic's influence on consumers' evaluation and decision-

(8)

making process, the second sees critics as mere predictions of the prospective audience. The main difference between the two approaches is the level of possible behavior changes of consumers through reviewers. Many authors have continued in this direction following this contribution to the field. Some have found evidence in favor of the influencer perspective (e.g., Boatwright et al., 2007; Ravid et al., 2006), some in favor of the predictor perspective (e.g., Gemser et al., 2007), and others in support of both views equally (e.g., Basuroy et al., 2003).

Shrum (1991) and Gemser et al. (2007) contrasted lowbrow and highbrow cultural goods regarding reviews' impacts. According to Shrum (1991), only positive reviews indicate a positive correlation with the performance of highbrow arts. The impact of reviews on lowbrow art is independent of the review's sentiment and only depends on the number of reviews. Combining the findings of Shrum (1991) and Eliashberg and Shugan (1997), Gemser et al. (2007) discovered that critics predict the audience's reaction to a mainstream (lowbrow) movie and influence the audience of an arthouse (highbrow) movie production.

Hennig-Thurau et al. (2012) studied the distinctions in a review's influence depending on the time. The authors found a stronger connection between reviews and a movie's short-term success than long-term success. With the increasing importance of online word-of-mouth (eWOM), Basuroy et al. (2019) compared the effects of expert reviews with those of consumers, discovering that experts' reviews yield a more significant correlation with performance. In contrast, Warne & Drake-Brooks (2016) stated that their "findings indicate that internet-based critics' opinions about productions […] are as respected and influential as newspaper-based critic's opinions" (p. 179).

(9)

Reddy et al. (1998) brought the research into the setting of the Broadway industry. The scholars studied the effects of informational sources, such as reviews and awards, and intrinsic show characteristics, such as the type and the cast. Both Simonoff & Ma (2003) and Nygren &

Simonoff (2014) build upon the work by Reddy et al. (1998) in studying the same variables. While Reddy et al. (1998) discovered the review in The New York Times to be more important than the one in the New York Daily News, Simonoff & Ma (2003) support an opposite view.

Additionally, some authors have concentrated primarily on a single factor in determining success, such as Boyle & Chiou (2009), studying the importance of the Tony Awards on the success of Broadway productions. Agreeing with the impact reviews have on the success of a production, Boorsma & van Maanen (2003) focused on the actual time of influence on the consumer. In their research, the authors found no evidence of reviews affecting the evaluation process before the ticket purchase. Meanwhile, their study uncovered that the consumer's opinion-making process post attendance is, indeed, impacted by critics.

Furthermore, Fraiberger et al. (2018) studied the importance of reputation to understand and determine factors ushering to success. The evidence that artists with a high-initial reputation were twice as likely to be part of exhibitions and that low-reputation artists commonly fail to be exhibited by prestigious institutions indicates that reputation influences the future success of an artistic good.

Interestingly, some scholars have taken a different road, challenging most previous findings. Walmsley (2011), an Arts & Entertainment professor at Leeds Metropolitan University in the UK, investigated the relationship between audiences and their emotions concerning the attendance of artistic production. Through his study, Walmsley (2011) found support in favor of

(10)

the hypothesis that emotions play a crucial role in the decision-making process of audience members.

Thinking back to the quotes mentioned at the beginning of this paper, sometimes the sentiment of newspaper reviews fails to align with future success, and sometimes it succeeds.

Even though some works included newspaper critics, past research has failed to shed light on newspaper reviews' effects on Broadway shows. Moreover, some works show contradictory findings on the importance of specific newspaper reviews (Reddy et al., 1998; Simonoff & Ma, 2003), introducing the question of whether all reviews have an equal impact on the success of a show. Did Wicked become a success because the review of USA Today has a more significant influence than The New York Times? To answer this question, this thesis proposes that, in addition to the review's sentiment, the newspaper's reputation plays a crucial role in determining the success of a Broadway show. The research question, hence, is as follows:

Does a newspaper outlet's reputation influence a review's effect on a Broadway musical? If so, which newspaper reviews are, hence, the most crucial ones?

This paper will have critical managerial implications in filling the research mentioned above the gap. If a newspaper with a more elevated reputation is more crucial in determining the success of a show, it might be beneficial to use quotes from those reviews for advertisement.

Furthermore, it may be possible for producers to seek a better professional relationship with critics from high reputation news outlets.

(11)

Due to the Broadway industry's tremendous economic value and size, an answer to this paper’s research question is highly relevant. During the 2018-19 season, 14,77 million people attended a Broadway show bringing in total revenue of over $1,8 billion. For years now, Broadway has been a crucial part of the economy of New York City, contributing to almost 97,000 jobs. In addition to ticket sales, Broadway brings in an additional $14,7 billion to the overall New York economy (The Broadway League). However, even though the business of Broadway, as a whole, seems to be highly successful, producers still struggle significantly with the factor of risk involved in every single production. The costs of producing a Broadway musical have statically increased with no apparent maximum. With the enormous amount of money at stake and the high risk involved, factors determining the possible success are crucial to all involved.

The method employed in this paper builds upon the core concept of Signaling Theory. In aiming to keep the information asymmetry between two parties to a minimum, Signaling Theory states that a signaler provides a piece of information in a signal to a receiver (Spence, 2002). The Signaler takes on the role of an insider, sending information that has since been unavailable to the outsiders, the receivers (Siering et al., 2018). Scholars have argued that the usefulness of a Signal to the receiver depends partly on the perceived quality of the signal source (Conelly et al., 2010). Scholars have borrowed aspects from the institutional theory and argued that reputation could drive signal quality (Deephouse, 2000; Coff, 2002).

This paper presents four main hypotheses based on the in-depth review of past works of scholars in this area. Following this, the methodology chapter will explain the conceptual model, the data collection process, and the method utilized throughout the rest of the paper. The subsequent chapter will provide a theoretical description of the analysis and its results.

(12)

Furthermore, the discussion will deliver a practical explanation of the results, provide theoretical and managerial implications drawn from the results, and limitations and propose suggestions for future research. Lastly, the paper ends with a conclusion.

(13)

2. Literature Review

2.1 Studies on the Effects of Critics on Market Performance

Eliashberg & Shugan (1997) were some of the first scholars to dive into the role of critics in the entertainment industry and advanced the research by differentiating between two different standpoints. According to the authors, critics can have a causal effect on a creative good, the so- called influencer perspective; they can also have a correlational impact, where critics take the role of predictors. While an influential critic changes the behavior of consumers, a predicting one merely foresees the audience's reactions. However, during their research, they only encountered some support for the latter, the predictor perspective. Eliashberg & Shugan examined the influence of reviews on the box office success each week after a movie's release date. The overall understanding of critics and reviews greatly benefitted from their work; however, it has since been criticized by multiple scholars. Boatwright et al. (2007) pointed out that the authors ignored the possibility of other explanations in their assumptions. Furthermore, later analysis contradicts Eliashberg & Shugan's study, only finding proof supporting the influencer perspective instead of the predictor one or both (Basuroy et al., 2003; Boatwright et al., 2007).

Basuroy et al. (2003) mirrored the approach taken by Eliashberg & Shugan (1997) in terms the data collection. Looking at eight weeks post movie release, the authors discovered that reviews of any sentiment, positive or negative, influence the box office numbers. This finding correlates with a dual perspective, indicating that reviewers can predict and influence.

Nevertheless, their results also demonstrated the influence of negative reviews decreasing over time, which correlates more with the influencer perspective.

(14)

According to Eliashberg & Shugan's (1997) theoretical framework, critics' opinions mirror the readers' tastes from the predictor's perspective. However, Boatwright et al. (1997) discovered that a review still yields an influence even with an apparent discrepancy between the reviewer's opinion and the audience. The authors assert that critics take the role of an information source, providing the readers with enough information to form their own opinion. Having acknowledged the issues of the Eliashberg & Shugan study, Boatwright et al. (2007) took a different direction.

According to them, results from past research failed to distinguish between the impact of the

"covariates—movie appeal and individual critic's impact" (Boatwright et al., 2007) on different movie genres. Hennig-Thurau et al. (2012) challenged the validity of the approach taken by Boatwright et al. (2007), as they examined deviations without including the consumers.

Most recently, Basuroy et al. (2019) compared the effects of consumers' electronic word of mouth with critics in the movie industry. Using time-varying data for the first time in this research field, the scholars found that reviews written by experts are of higher importance than audience evaluations, both in valence and volume. Valence refers to the quality of reviews, whether positive, negative, or neutral. Interestingly enough, the paper provides evidence that the influence of expert reviews increases with a higher resemblance between the reviewer's and the audience's opinions.

While many authors focused on the impact of reviews on the performance or sales of production, Chen et al. (2012) diverted the perspective by studying the value of third-party reviews for investors. The authors use a third-party review as a synonym for a critic's or expert's review. The author uncovered that critics influence investors during the reintroduction stage to predict future success. However, the review's sentiment is not necessarily directly linked to its

(15)

influence. Even a negative review might positively affect an inventory, but only if the review is still more favorable than a previous one.

This shows that numerous scholars have uncovered support for a positive correlation between positive expert reviews and the box office success of movies (e.g., Eliashberg & Shugan, 1997; Hennig-Thurau et al., 2012). It is crucial to mention that, while many studies have concentrated on the movie industry, the similarity within the creative industries allows for applying research findings to the theater industry.

2.2 Differentiating between Highbrow and Lowbrow Culture

Shrum (1991) studied the differences in the effect of reviews on highbrow and lowbrow art.

According to the author, positive expert reviews, such as critics, positively correlate with the success of highbrow art performances. The author defined highbrow art to be "expected by cultural reasoning" (Shrum, 1991, p. 356), as, for example, the theater or opera. As for lowbrow or popular culture, the mere attention of critics overshadows the actual evaluation of the critic.

Hence, the valence plays a more central role than the review's sentiment for lowbrow art.

Overall, the author argues that even a negative expert review is more beneficial than none. This statement supports Hirsch's (1972) work, stating that attention (volume) generally already positively supports performance.

In line with the work of Shrum (1991), Gemser et al. (2007) implied that the influence of reviews differs for the audiences of art-house movies and mainstream movies. The authors concentrated on the Dutch movie industry and analyzed box office data, newspaper reviews, movie types, and other control variables, such as star casting and the number of screens. Through

(16)

their regression analysis, they discovered empirical evidence supporting their original hypotheses. Movie reviews foreshadow the audience of mainstream movies, whereas they influence the audience of arthouse movies. Additionally, they uncovered the valiance and size of reviews to play a crucial role within the first week's post-release.

Both Shrum (1991) and Gemser et al. (2007) took a different approach in investigating reviews' performance effects compared to the authors in the previous section. They contrasted two categories of entertainment goods, high and lowbrow. According to Shrum (1991), as the theater is generally considered a highbrow art, the quality of the review is of higher importance than its valence. Drawing a parallel from Gemser et al. (2007) to the theater industry is more complex, as the authors distinguished between art-house and mainstream movies. While many argue that Broadway musicals are mainstream, no empirical evidence reinforces this. However, these two papers demonstrate that the influence of reviews varies depending on the entertainment good in question.

2.3 Studies in the Theater Context

So far, the papers discussed primarily focused on the effect of reviews on movies, partially resulting from only a handful of scholars centering their research solely on the theater industry.

One of the few studies assessing the theater industry is Hirschman and Pieros (1995), who focused on the indicators of success, comparing Broadway plays with movies. While the authors uncovered a negative correlation between reviews and a movie's box office performance, they saw no correlation between Broadway reviews and performance. However, their study

(17)

demonstrated tremendous limitations, as the data collected consisted of only ten plays and ten movies.

Reddy et al. (1998) investigated the factors determining the success of Broadway musicals and plays. With a primary focus on information sources used in the selection process of consumers, the immediate research question was the impact of theater reviews on the success of a show. More closely, Reddy et al. (1998) utilized the measure of attendance numbers and number of performances as a guideline to measure success. Aside from this, the authors looked at other possible influences, such as the number of previews, ticket prices, advertising, and intrinsic features, such as the type of show, the cast featured, and the opening time. Through a linear regression, the scholars found support for the crucial influence newspaper critics, in general, have on a show's success. However, even more, striking is that their research indicated that the review from The New York Times yielded a more significant influence than the ones in the Daily News and New York Post. Reddy et al. explained this to stem from two possible aspects:

The New York Times reviews tended to be longer and more detailed in text than the other reviews. The newspaper's demography possibly resembled the typical Broadway audience better than others. Unfortunately, the author neglected to discuss possible explanations for this finding further. While the authors acknowledge that expert reviewers may act as influencers and predictors, their results support Eliashberg & Shugan's (1997) findings supporting the influencer perspective. Interestingly, as for the limitations of their study, Reddy et al. (1998) recognized that the same author wrote many reviews for one of the selected newspaper outlets.

Simonoff & Ma (2003) criticized Reddy et al. (1998) relatively old data selection (1980- 82). Using a similar approach as their predecessors, Simonoff & Ma's data consisted of Broadway

(18)

shows within three years. Based on the previous findings of Reddy et al. (1998), the newspaper outlets utilized in this study were The New York Times and the New York Daily News. Through their quantitative research, the scholars found evidence that the type of show (e.g., musical, play) influences longevity, where musicals tend to have more prolonged runs than plays. While Reddy et al. (1998) found The New York Times to correlate with success, Simonoff & Ma (2003) discovered contradicting evidence. Moreover, the authors uncovered reviews in the New York Daily News to influence the longevity of a theatrical show.

Compared to Simonoff & Ma (2003), Nygren & Simonoff (2007) took a different approach in employing the Cox proportional hazards model technique. According to the authors, this technique allows for testing the adequacy of the model through a graphical manner and hypothesis testing. However, similar variables were employed, such as longevity, critic's reviews, type of show, and acknowledgment at the Tony Awards. While the scholars used a before- uncommon method, the findings directly matched the ones of Simonoff & Ma (2003).

Boorsma & van Maanen (2003) noticed a gap in research in terms of the point in time of the effect of reviews on consumers. Though their findings supported past research on the importance of critics on a theatrical performance's success, the authors were interested in whether the influence occurs prior to the performance or after that. The scholars found that newspaper reviews influence the audience through their case study on the Dutch production Mijn Elektra. Contradictory to the common belief, the authors found that newspaper reviews have no impact on the decision-making process of attending a performance but on the perception-making process after the performance.

(19)

Warne & Drake-Brooks (2016) focused on how information sources influence a consumer's decision-making process for a theater ticket purchase. Since the newspaper industry has changed drastically between earlier studies, such as the one by Reddy et al. (1998), Warne &

Drake-Brooks (2016) argue that, as a result, online sources have stepped up and gained influence.

According to Warne (2018), the number of cities with full-time newspaper theater critics is only 11 in the US. Their study showed that blogs and social media influence the willingness to buy a ticket, similar to newspapers. Additionally, according to the participants, reviews published on blogs are as professional as newspaper reviews.

An article in The Economist (2016) allayed the ingredients of a successful Broadway show with the example of the hit show Hamilton: An American Musical. According to this article, two approaches to business success have shown success in the past. On the one hand, success might be guaranteed by adapting previously thriving movies into stage productions or having a famous actor or actress take a leading role. The latter method lifts the production's likelihood of selling out its first week from 21% to 59% with a famous person and 92% with an A-lister. In contrast to previous research, the paper asserts that critics are wrongfully believed to influence the success of a show, as "a musical with a rave review in The New York Times is less than six percentage points more likely to sell out in a given week than one with a neutral review" (The Economist, 2016).

Cited in The Economist article is a paper by Kulmatitskiy et al. (2015) studying the trends of Broadway shows and their survival in the industry. Building upon Simonoff & Ma (2003) and Nygren & Simonoff (2007), the authors not only focus on more recent data (2000-2009) but also apply a combination of a proportional hazard model, logistic and linear regression. The authors

(20)

put the Tony Awards as the most significant industry recognition center stage and divide a season into pre- and post-award seasons. As the scholars argue that the audiences demonstrate substantial differences within a season, the season was further divided into pre-season and pre- nominations periods. Their results indicated that written reviews, such as newspaper reviews, are only influential during a specific period. In this period, between the announcement of Tony Award nominations and the award ceremony, reviews in the Daily News showed a positive correlation with success. This stands contrary to the findings of Reddy et al. (1998) and supports Simonoff & Ma (2003).

Overall, there is overwhelming support that there is a correlation between expert reviews and performance. However, past research has found contradicting results regarding the importance of specific newspapers in the theater industry. From this, the question arises if the influence of newspaper reviews on a show's performance differs depending on their outlet.

Suárez-Vázquez (2011), who studied the effect of star power and critical reviews on movie audiences, emphasizes the importance of this topic by stating that: "it would be […] interesting to know whether their effect depends on the medium that they appear in "(Suárez-Vázquez, 2011, p. 132). Furthermore, the follow-up question arises if there are factors influencing the effect a newspaper review has on the performance of a theatrical production. Is there a moderating effect influencing the power a newspaper review has on the success of a Broadway show?

(21)

2.4 Ranking in Expert Reviews and Journalism

Critical evaluations act as the mediating party between the producers of cultural goods and the prospective consumers. According to Bourdieu (1980), expert criticism provides a form of legitimacy to the product in question as the critics are believed to have expert knowledge. With the rise of the World Wide Web, new forms of critical evaluations have emerged, for example, Internet reviews. Verboord (2009) studied the different effects of critics “of varying degrees of institutionalization” (p. 623). The author bases the paper on the assumption that the importance of critics shows a positive correlation with the level of institutionalization. Interestingly enough, Verboord used the example of The New York Times, proposing that a reviewer of this newspaper holds a more significant influence on the reviewed products. Additionally, the author brought in the concept of omnivorous taste. Here, results showed that critics evaluating various genres align with a higher form of believability.

The author Kammer (2015) puts online amateur reviews at the center of attention. In line with other scholars, Kammer argues that the impact of reviews can be ranked according to the legitimacy gained through the publishing outlet. While expert reviews are perceived to present an objective evaluation, amateurs often lack this perception.

In general, these studies illustrate that consumers put different levels of importance on expert reviews depending on the perceived legitimacy of the evaluation.

2.5 Reputation Influencing Success

Fraiberger et al. (2018) pursued quantifying the importance of the reputation of and within a network to gain a deeper understanding of success in the art world. According to the authors,

(22)

artists and institutions can access the necessary resources through reputation and networks, which lead to success. Their extensive data set comprised about half a million global exhibitions over 36 years. In their analysis, the scholars created a spiderweb of connections between high- and low-prestige institutions and artists. Overall, Fraiberger et al. (2018) discovered that artists with an above-average initial reputation exhibited twice as many occasions as artists with a lower initial reputation. Furthermore, the results showed that artists with a low reputation struggle to move up the ladder to a prestigious institution. Hence, evidence of the impact of reputation on future success seems to be apparent.

As previously mentioned, not all newspapers seem to have an equal effect on the performance of a theatrical performance. Moreover, studies have shown legitimacy's impact on the importance of a review. As reputation plays a crucial role in the success of a cultural product and can act as a method of legitimization, the following hypothesis is proposed:

H1: The reputation of the newspaper outlet has a positive effect on the influence of a review published by the given outlet on the performance of a Broadway show.

2.6 Signaling Theory

It is hardly ever possible to directly observe all the aspects of interest in any communication.

Instead, signals commonly provide information on hidden features that better understand a communication situation. However, signals are also prone to be either misunderstood or purposefully deceived.

(23)

The Signaling Theory was initially developed in the field of economics (Spence, 1973) and later adopted in the biology field (Zahavi, 1975) to minimize the asymmetry of information between two involved parties (Spence, 2002). The actor with more information creates a signal intended to “communicate” and “carry information” (p. 434, Spence, 2002) to the actor with less information. Scholars have commonly used this theory in connection with the principle-agent theory, where the principle holds more information than the agent.

The key concepts within Signaling Theory are the signaler, the signal, and the receiver.

The following paragraphs will explain it all in more detail. The main focus of Signaling Theory lies with the signalers, the creators of signals themselves. Signalers are experts with in-depth knowledge of the good or service in question who have gained insider information that has not been available to the broad public yet (Siering et al., 2018). While the information in question may be of positive or negative origin, it is believed to be of value and interest to others (Conelly et al., 2010).

The information gained by the signaler is the signal itself. Spence (1973) distinguished between two types of signals, assessment signals and conventional signals, differing in the level of perceived reliability. Assessment signals are highly reliable due to their mode of production, as it directly requires the possession of quality. As for conventional signals, there is no inherent need to possess the quality. According to Siering & Muntermann (2013), both types of signals can be found in information signals. Research in online reviews has differentiated between signals related to the review content (review-related signals) and the review platform (reviewer-related signals). While review-related signals are relatively easy to influence, reviewer-related signals come with a higher cost as they commonly require continuous activity throughout time and

(24)

verification. Additionally, reviewers are usually unable to manipulate the latter signals (Siering et al., 2018).

However, not all signals are equal in quality and usefulness to the receiver. Conelly et al.

(2010) coined the term signal reliability (also called credibility). It combines two crucial aspects of the effectiveness of a signal, the match of the desired signal quality of the receiver and the actual quality of the signal and the perceived honesty of the signaler. Past scholars have studied various ways of identifying the quality of signals. A promising approach here seems to merge the institutional theory, which argues that legitimacy is the core ingredient for a firm’s survival, with the Signaling Theory (Certo, 2003). According to Deephouse (2000) and Coff (2002), a firm can increase its signal quality by increasing its perceived reputation.

A signaler sends information in the form of a signal, which is then obtained by the receiver. A receiver seeks out information to fill a personal knowledge gap. Hence, the act is rooted in personal motivation. Consumers search for information before purchase to minimize the risk of dissatisfaction.

Signaling Theory has had a tremendous impact on research across various fields. Designed initially to resolve the asymmetry of information between actors, its core concept states that a signaler transmits a piece of information in the form of a signal to a receiver. While Signaling Theory has been highly researched, especially in online reviews and user-generated content, researchers have failed to look at the topic of newspaper reviews as signals in the theater industry. Siering et al. (2018) studied reviewer-related signals as social platforms or user-specific information. However, this paper takes a different approach in identifying reviewer-related signals as the reputation of a newspaper outlet in which the review is published. Borrowing the

(25)

vital elements from institutional theory, this paper proposes that the information source’s reputation behaves as a form of signal credibility. As a result, this paper argues the following:

H2: The newspaper outlet's reputation moderates the influence a review published by the given outlet has on the financial performance of a Broadway show.

2.7 Studies on Moderating Effects

Hennig-Thurau et al. (2012) distinguished between a movie's quality evaluation of the audience and reviewers. Furthermore, they studied the effects of different periods. Their study supported the more significant influence of reviews on short-term rather than long-term performance.

Moreover, they dove into possible moderating effects. For the short-term performance, they discovered advertising and buzz to be vital. As for long-term success, both celebrity casting and sequels positively moderate the results, while a movie's genre has a negative moderating effect.

While scholars studied various aspects, such as reviews, genre, or type, and start casting, influencing the success of a cultural good, Desai & Basuroy (2005) saw a lack in past research on the combined effect of these factors. Using 275 movies over two periods, the author empirically tested the moderating effects of the genre, star power, and reviews on performance. Concerning reviews, the authors found that the quantity has a negligible impact on the performance of a movie with a standard and familiar genre. In contrast, reviews for movies in a less familiar genre have a significant favorable influence. The familiarity of movie genres was based on the number of releases of movies in a particular genre in a given year.

(26)

As authors include other factors in addition to critical reviews in their studies, the question arises if the previously proposed moderating effect of the reputation of an information source can become negligible. According to Signaling Theory, a receiver uses signals to fill an information gap. Hence, it could be claimed that if a receiver holds previous knowledge, the signal loses its value on the receiver side of the signal. Broadway houses either new productions or revivals, which are previously produced shows. As a result, prospective audiences of revivals may not need to rely on expert reviews as extensively as they might for new productions.

Additionally, audiences may have some prior knowledge if a show is sometimes based on existing content, even for new theatrical productions. Examples of this are musical adaptations of movies or books, for example, the musical Lion King. Moreover, so-called Jukebox musicals, which use existing and established music as a show's framework, have become more popular.

Within Jukebox musicals, one finds two types of shows. The storyline might be entirely original, created to fit the existent music, or based on artists' or bands' lives. Due to consumers' prior knowledge, the following additional hypotheses are suggested:

H3: The moderating effect of the reputation of a newspaper outlet on the influence of newspaper reviews on the performance of a Broadway show diminishes if the show is a revival.

H4: The moderating effect of the reputation of a newspaper outlet on the influence of newspaper reviews on the performance of a Broadway show diminishes if a show is based on previously known content.

(27)

2.8 Studies on Measurement of Success

Previous papers have looked at the success of a Broadway show in terms of its longevity or financial performance. However, it is vital to keep in mind various methods of measuring success.

The following paragraph discusses two alternative ways.

In 2010, the Independent Theatre Council, Society of London Theatre, and Theatrical Management Association (both UK) challenged the previously mentioned studies by taking a very different approach to measuring a show's success and quality. Charlotte Jones, the chief executive of The Independent Theatre Council, noticed a lack of ways to determine a show's success that correlate with the original intention of the play's or musical's composition. According to Jones, the main reason for creating a theater piece is to evoke the emotional responses of audience members. With this in mind, Jones created their so-called "well-being" toolkit, which included post-show surveys asking the audience about their feelings during the performance and whether or not they noticed the time passing while experiencing the show. According to Alistair Smith (2010a), an editor of The Stage UK, the results of this come in a graph showing the broadness of emotions covered by a show, stating that "the larger the area, the more successful the show" (Smith, 2010a).

Following this, Dr. Ben Walmsley (2011), an Arts & Entertainment professor at Leeds Metropolitan University in the UK, studied the audience's emotions during theater productions from a psychological perspective. The author based his work on a previously published study by Jeremy Pincus, who argued that no research had proven the "behavioral and emotional meaning of unmet needs" (Pincus, 2004). With this in mind, Walmsley contradicts previous assumptions that emotional experiences have negligible importance in satisfying deep-set values, which are

(28)

the primary motivation behind theater attendance. From a more anthropological point of view, Walmsley analyzed motivation drivers, spiritual, emotional, intellectual, and social. Through responsive interviews and participant observation, Walmsley found emotions to be a highly compelling reason people decided to experience a performance at a theater. Additionally, the author's results showed that most audience members look for pure entertainment in addition to a challenge to escape their everyday lives. As a result, Warne implies that a show's production and creative team benefits from understanding the audience's relationship and emotions.

(29)

3. Methodology

The main goal of this paper is to find empirical support for the moderating effect of the reputation of newspaper outlets on the influence of reviews on the performance of a theatrical performance. More specifically, this paper studies this topic in the research field of Broadway musicals. This chapter begins with an overview of the conceptual model based on the previously introduced Signaling Theory. After that, the overall research design will be explained in more detail, and this chapter concludes by shedding light on the utilized variables.

3.1 Conceptual Model

After carefully reviewing past literature, a conceptual model was designed. Signaling Theory provides a perfect building block for the conceptual model of this paper. As discussed earlier, the core concept of Signaling Theory is as follows: the signaler sends information as a signal to a receiver. In the context of newspaper reviews of Broadway shows, the review's author takes on the role of the signaler. In the Broadway industry, reviewers are commonly invited to rehearsals and previews ahead of the production's official opening. This enables the reviewers to gain insider information unavailable to others. Signalers, then, can share this information with others through articles, including reviews. The review becomes a signal as soon as it is made available to the general public, either through the newspaper's printed form or digitally. Interested parties, the receivers, may then seek out this information to fill a previous knowledge gap.

Most research focused on the effect of reviews on the performance of creative goods has failed to study a possible moderating effect. Scholars have found the importance of reviews and

(30)

reputation on performance but never in combination. Hence, this paper proposes that newspaper reviews influence the success of a Broadway musical based on its reputation. In theory, the reviews’ sentiments stand as the independent variable and the net grosses as the dependent variable. This paper decided to include the number of performances as a mediator affecting a show's financial performance. As reputation has a proposed effect on the influence of the reviews on the performance but has no direct impact on either the review or the performance itself, reputation is suggested to have a moderating effect. The following graph shows a simplification of the proposed theoretical framework.

Figure 1: Conceptual Model Sentiment of

Newspaper Review

Net Grosses Number of

Performances Reputation of

Newspaper Outlet

(31)

3.2 Research Design

As for the research topic of this paper, it was decided to focus on quantitative research. It follows the philosophical approach of positivism, as the data collected can be analyzed objectively.

Before going into further detail about the data collection process, it is crucial to look at the overall restraints of this research.

A Broadway show is not simply a theatrical production in New York City. To be considered a Broadway show, several additional requirements must be fulfilled. In addition to the location in Midtown Manhattan, the theater generally needs to seat at least 500 people. Interestingly enough, out of the 41 official Broadway theaters, only three are located on the actual Broadway street (Playbill Staff, 2021).

This research considers Broadway shows with an official Broadway opening between January 2007 and December 2019. Hence, the data does not reflect out-of-town runs and previews. Broadway houses a great variety of shows; however, the primary focus of this research is Broadway musicals. Productions of plays on Broadway commonly have a definite length of performance set in advance. According to an article on Playbill.com, it is a highly profitable option for producers of plays to “secure the services of a name star and nail him or her down for a run of anywhere from 10 to 20 weeks” (Simonson, 2007). A limitation of performances creates exclusivity for the audience and increases demand. As this paper measures success based on financial grosses and the number of performances and plays commonly have restricted numbers, the decision to solely focus on musicals was made. Therefore, the data disregards all plays, variety, dance shows, and concerts.

(32)

Furthermore, all musicals that opened with a set closing date, also called limited-run shows, are left out of this research. As for the time, it was decided to be as current as possible:

However, due to the tremendous effects the COVID-19 pandemic has on the entertainment industry as a whole and, therefore, on Broadway, any shows having opened in the affected season of 2020-21 and 2021-22 are dismissed. This decision is based on the fact that even after the national- and state-wide lockdowns, the global pandemic regularly shut down theaters for about 18 months (Broadway Throws Open Its Doors after Longest Shutdown, 2021). While some productions postponed their opening to a later season, many shows were dismissed or closed during this period. However, some shows survived the period and are up and running again (Geier, 2021). Hence, excluding musicals with their official opening before the pandemic will allow for representative and comparable results.

Considering the abovementioned aspects, 125 musicals tick all these boxes and are part of this research. The official opening dates range from March 2007 to December 2019, with an average run of 500 performances. While eight shows reopened after the national lockdown, the performance numbers and grosses will exclude the time frame after the reopening after being closed for 18 months. Out of those shows, six musicals continue to run on Broadway (as of April 2022).

3.3 Primary Data Collection

There are numerous newspaper outlets regularly publishing reviews of Broadway musicals. An online survey was distributed among a convenience sample to create a newspaper selection utilized in this research and provide the newspaper outlets with a reputation ranking. For a

(33)

valuable result, the overwhelming number of participants either currently live in New York City or have in the past. Furthermore, all respondents had to have seen at least one Broadway musical before and have previous knowledge of the theater industry. These restrictions will allow for probable results. The convenience sample included current and past students of Arts &

Entertainment Management students from Pace University in New York City and professionals working in theater administration or entertainment marketing industry in New York City. During the survey, respondents are asked to rank a section of 10 newspapers that regularly publish Broadway reviews. The results of this survey ranking act as a guideline in selecting newspapers used in the further analysis. It must be noted that the respondents were not provided any material or financial incentive for their participation in this survey. The entire survey questionnaire is provided in appendix 8.1 of this paper.

3.4 Variables

This paper analyzes the moderating effect of the reputation of the source of a newspaper review using multiple regressions. As outlined in the conceptual model, the three primary variables at the center are newspaper reviews, a newspaper's reputation, and the performance of a show.

Following this, all variables are explained in more detail.

3.4.1 Newspaper Reviews’ Sentiment

A Broadway musical is a so-called experimental good coined by a relatively high risk for the consumer. A prospective consumer cannot try the product during the decision-making process or return if unsatisfied. Furthermore, the consumer's value creation occurs during or after

(34)

consumption. However, information sources can help consumers minimize their risk ahead of the purchase (Nygren & Simonoff, 2007). In the Broadway setting, these information sources are very commonly theater reviews. Drawing a parallel to the concept of the Signaling Theory, reviews take on the role of the signal. As previously mentioned, signalers, in this case, the reviewers, provide insider information to outsiders seeking it.

Information sources may be classified as experiential or objective. A critical review may allow the reader to imagine the feeling of experiencing the show themselves (Cooper-Martin, 1992; Faber and O'Guinn, 1984). In addition to this, information sources may include valuable information and objective facts. The perfect example of objective information is a plot summary, which most reviews have. Moreover, background information on the cast and leading team might be part of a review (Reddy et al., 1998).

Critical evaluations are seen as credible information sources and can persuade readers.

Due to a critic's experience and knowledge, consumers attribute credibility to the critic. The authors of reviews are also believed to have researched all required information without any previous bias (Reddy et al., 1998). Additionally, the credibility of a review might also stem from the credibility of the platform it is published on, in the case of a newspaper outlet's newspaper review. This combines the concept of the Signaling Theory with the institutional theory.

For the means of this paper, while there are a variety of newspapers regularly publishing reviews of Broadway shows, it was decided to focus on six newspaper outlets. The selection decision was based on the online survey results mentioned above. The chosen newspapers are as follows: The New York Times, Variety, The Guardian, New York Daily News, Vulture, and Hollywood Reporter.

(35)

All reviews used in this research have been retrieved from the newspaper's archive.

Unfortunately, it was impossible to collect reviews for all 125 shows from all six newspaper outlets, as some newspaper outlets only reviewed a selection of shows. The New York Times and The Hollywood Reporter only lacked two and five expert reviews. While the New York Daily News

and Variety neglected to review 22 and 17 musicals, respectively, the reviews of 35 shows were missing from Vulture. Unfortunately, during this research, it was only possible to find the reviews of about 40% of the musicals from The Guardian. The following table provides an overview of the number of reviews collected for each newspaper outlet.

Newspaper Outlets Number of Reviews

Daily News 109

Hollywood Reporter 126

New York Times 129

The Guardian 53

Variety 114

Vulture 96

Table 1: Number of Reviews per Newspaper Outlet

Unfortunately, only one newspaper, The Guardian, includes a numerical ranking with theater reviews. Due to this lack of comparable rankings, some studies that have analyzed newspaper reviews primarily focused on the valence of reviews (Basuroy et al., 2019; Shrum, 1991); others have used independent judges to determine the review's sentiment (e.g., Reddy et al., 1998;

Simonoff & Ma, 2003). Two industry experts evaluated a selection of newspaper reviews and

(36)

gave a 1-5 rating. Reddy et al. (1998) even attributed their use of only two post-graduate students as judges as one of their study's main limitations. As for the means of this paper and the fact that the review rankings or scores themselves are not the primary focus, all reviews’ text will be evaluated by algorithmic textual LIWC sentiment analysis. Furthermore, this will allow for an unbiased evaluation based on textual elements. The results of this analysis will provide whether the critic has a positive, negative, or neutral tone, which, in turn, will be used to define a positive, neutral or negative review further on. Based on this sentiment analysis, the reviews will be allocated with sentiment rankings between 1-100, whereby a score under 50 equals a rather negative sentiment, and above 50 is a more positive sentiment.

However, as theater critics commonly draw from creative writing aspects, such as metaphors, humor, and irony, sole sentiment analysis may not be reliable. Therefore, the outcome of the sentiment analysis will be combined with the review ratings provided by the theater website BroadwayWorld.com (BW). This website's influence can be demonstrated at the hand of its monthly visitor rates. According to Alexa, the website registered around 5.5 million monthly visitors in 2018 (n.d.). After a show's opening day, BW compiles a list of reviews, ranking each review based on its tone. This ranking ranges from one to ten, where a score of one refers to a highly negative sentiment, and ten refers to a highly positive sentiment. While the theater website included the majority of existing newspaper reviews in its ranking, it neglected to provide a ranking for about 20% of reviews. However, it is crucial to mention that these rankings are not complete. As above-mentioned, not all shows were reviewed by experts from all newspapers.

Hence, there are no rankings from BW for those, either. Additionally, the BW rankings from other reviews were missing. This paper's analysis section uses an average of the sentiment analysis

(37)

results and the rankings provided by BroadwayWorld.com. As for these cases, only the sentiment analysis results are employed in the testing process.

3.4.2 Reputation

At this point, it has been established that reviews are signals delivered to information seekers and influence the performance of a theatrical production. While a tremendous amount of information circles the world, only a minor fracture of which is consumed. As it is for every good or service struggling with oversupply, certain products need to stick out to be chosen. As for signals, receivers desire to select the signal with the highest level of usefulness to reduce waste and search costs. Important factors of a signal’s usefulness are perceived quality and the signalers expected honesty. Credibility and reliability play a crucial part in both of these factors. Combining Signaling Theory with Institutional Theory, scholars have argued that reputation can help increase the credibility of the signal and the reliability of the signal source. This supports the previously made proposal that the reputation of a newspaper outlet (signal source) has a moderating effect a review has on the performance of a show.

The most common approach to measuring the reputation of newspapers is looking at the circulation and readership numbers. However, as these numbers would not specifically focus on the arts & entertainment sections of newspapers and, as a result, lead to a biased ranking, this method was not employed. While there are numerous ways of measuring the reputation of a newspaper outlet, it was decided to rely on the results of the primary data collection. This will allow for a ranking precisely mirroring the perceived reputation of readers interested in the theater industry.

(38)

During the online survey, respondents were asked to rank a selection of ten newspapers, all of which frequently publish Broadway reviews, from one to ten. The list of newspaper outlets was provided to the respondents. Thereafter, an overall ranking was calculated as the average of the respondent’s orders. The following graph (Fig. 2) shows the survey results; the newspaper outlets are ranked according to their reputation.

Figure 2: Newspaper Ranking (as per survey)

Furthermore, the ranking played a crucial role in the selection of the newspaper outlets used throughout the rest of this paper. In order to receive representative results, it was decided to pick newspapers from both ends of the ranking as well as two from the middle. The New York Times and Variety lead the field with an average ranking of 8.44 and 6.89. The Guardian (567)

and the New York Daily News (5.40) are in the middle. Lastly, Vulture and the Hollywood Reporter bring up the rear with rankings of 5.19 and 4.90.

Vulture The

Guardian

Entertainme nt Weekly

New York Post The New

York Times

Variety Vogue Associated

Press

New York Daily News

Hollywood Reporter

(39)

3.4.3 Performance

According to New York Show Tickets Inc. (2021), Hamilton: An American Musical is the best musical of all time. According to playbill.com, the number one top-grossing Broadway musical is Disney's The Lion King (The Top 10 Highest-Grossing Broadway Shows of All Time, 2022).

According to The Broadway League (n.d), the longest-running Broadway musical is The Phantom of the Opera. Most industry insiders and experts in the field will not hesitate to refer to any

musicals mentioned earlier as successful Broadway shows. However, these three musicals earned their respective "title" based on different main foci, such as the overall box office profit or the length of the run. According to the Cambridge Dictionary, the word success can be defined with the following two definitions: "the achieving of the results wanted or hoped for" and "something that achieves positive results" (Success [Def. 1 & 2]. (n.d.). However, these two definitions leave much room for interpretation, as the results one hopes to achieve can be highly subjective and may vary from person to person. As a result, the same thing might have achieved success in one person's eyes but not in the eyes of another. The example provided above perfectly ties in with this by showing that the word "successful" has numerous ways of interpretations.

One could argue that a Broadway musical's first level of success is the moment of recoupment. In the theater lingo, if a production recoups, it has made enough revenues to regain the entire initial investment. Technically, a show only starts to make a real profit after that. There is a common trend in the creative industries, where a few big players bring in the majority of revenue, and many smaller players drop out. Drawing a parallel to the Broadway industry, only a few shows succeed in running for multiple years and, even less, achieve hit status, staying on Broadway for a decade.

(40)

While numerous ways of measuring a Broadway show's success have been investigated, most scholars have treated success and financial performance as equals (e.g., Reddy, 1998). This decision of selecting revenues as a determinant of success is assumably based on the fact that Broadway grosses are public information that can be easily obtained. However, other interesting information, such as weekly expenses or production costs, is almost impossible to acquire.

Ticket prices of Broadway shows vary greatly, with an average of $122,73 during the Broadway season 18-19 (Statista, 2022). Nevertheless, in June 2019, the best seat in the Richard Rogers Theater came with a hefty price of $849. Still, economists argued that even these prices were underpricing due to the extreme demand for the Broadway musical Hamilton: An American Musical. The premium seats and the average admission prices to see the musical about the

founding father Alexander Hamilton ranged from $270-$310 that year (Passy, 2019). These ticket prices will result in higher weekly revenue and, hence, higher financial numbers. To account for this divergence, the longevity of a Broadway musical will be taken into account in addition to the revenue. In this paper, a show's longevity equals the number of performances played.

Overall, the longevity ranged from one mere performance for Glory Days to 3.746 performances performed by the Broadway cast of Book of Mormon. The average longevity in terms of performance numbers of the selected shows lies at approximately 500 performances.

The highest and lowest net grosses are in line with the number of performances, with Glory Days on the lowest end of the spectrum for financial performance, reaching only $ 1.209.478, and on the highest end Book of Mormon, with $ 659.604.381. However, the lack of perfect congruence between the number of performances and the net grosses earned can be shown in the example of Hamilton: An American Musical. While the musical has been performed on Broadway 1.942

(41)

times, which is about 50% less compared to Book of Mormon, the show has almost caught up with the net grosses of the set musical. Every theater houses a different number of people, and the ticket prices vary across Broadway musicals, which explains this phenomenon. Furthermore, this supports the decision to include the longevity of shows and financial performance.

All required data, such as opening and closing dates, number of performances, and overall grosses, were retrieved from BroadwayWorld.com, one of the leading information sources on the worldwide Broadway industry. Having one source for this information provides a valuable framework. The list of shows, including its official opening and closing dates, the number of performances, and the net grosses, is included in appendix 9.2.

3.4.4 Production Type

While this paper suggests that reviews' influence on performance is dependent on the reputation of the reviews' outlet, it also proposes the possibility of this effect diminishing under certain conditions. Aside from expert reviews, there are other sources from which prospective consumers can gain information. While some musicals open on Broadway for the first time, others have had a previous run on Broadway, such as the musical Cats, which was brought back to the Broadway stage in 2016, ten years after its first production closed. The theater industry refers to shows returning to the Broadway stage as revivals. Even though a revival is most likely to star an entirely different cast and commonly also a new creative team, some consumers will forgo the active information search for the new production based on their already exciting knowledge from a previous version of the show. After all, the plot and the music will remain the same, with possible minor changes here and there. Therefore, the data used for this research

(42)

include revivals. The data list indicates whether a musical is shown for the first time (New Production, NP) or brought to the stage again (Revival, R).

3.4.5 Content Type

Additionally, the creators of musicals may base their show on already existing content, either in the form of a storyline or music. Inspiration may be drawn from historical events, books, or movies. Moulin Rouge, based on the 2001 movie (Moulin Rouge! The Musical, 2021), and Hadestown, based on the ancient Greek mythology (Hadestown | Official Site, n.d.), are examples

of the data used in this research. Furthermore, Jukebox musicals use well-known music and either create a story around the music or tell the story of the band or artist. Beautiful: The Carol King Musical tells the story of the singer and songwriter Carol King and features some of their most famous songs. The musicals are categorized as having original content or based on previous content in terms of the research data.

3.4.6 Control Variables

Control variables are incorporated in the analysis section to allow for representative results. For the means of this paper, it was decided on the following control variables: duration of the show, previews before the official opening, and the theater house. As for the length of the performance, the data is provided in minutes and does not include intermissions.

(43)

3.4.7 Variable Summary

The following table summarizes all utilized variables.

Table 2: Variable Summary

Variable Name Description

PERF Number of performances played up to the end of the production’s run or up to the begin of the national lockdown in the US.

(Continuous Variable, Mediator)

GROSS Total number of net grosses made up to the end of the production’s run or up to the begin of the national lockdown in the US.

(Continuous, Dependent Variable)

SENT Average of textual sentiment analysis (LIWC) of expert review and the BroadwayWorld.com rating of the review measured on the scale of 0 to 100. (Continuous, Independent Variable)

SENT_AN Tone results of textual sentiment analysis (LIWC) of expert review on a scale of 0 to 100. (Continuous Variable)

BW_RANK BroadwayWorld.com rating of the review measured on a scale of 0 to 100.

(Continuous Variable)

DUR Duration of the production measured in minutes excluding any

intermissions. (Continuous, Control Variable)

THEATER Name of the theater of the production. (Categorical, Control Variable) PREV Number of previews prior to the official opening date.

(Continuous, Control Variable)

(44)

REP Average survey results measured on a scale of 0 to 10. (Continuous Variable, Moderator)

CONT Dummy Variable, where original content is indicated by “1” and not original content by “0”. (Moderator)

ProTyp Dummy Variable, where a new production is indicated by “1” and a revival by “0”. (Moderator)

The Power of the Press

Full text

MSc Business Administration:

Entrepreneurship & Management in the Creative Industries