To fit in, or to stand out? : the moderating role of status on the relation between (non-)conformity and rap song performance

(1)

To Fit in, or to Stand out?

The Moderating Role of Status on the Relation Between (non-)Conformity and

Rap Song Performance

Student: Jeroen Hillebrand

Student ID: 10549099

Date: 21-06-2018

Master Thesis: MSc. Business Administration

Track: Entrepreneurship and Management in the Creative Industries

(2)

2

Statement of Originality

This document is written by Jeroen Hillebrand who declares to take full responsibility for the

contents of this document. I declare that the 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

(3)

3

Abstract

Actors have to choose between conforming to and differentiating from their category’s standards. Past research found that one’s status-level influences this decision, however, this study goes a step further by examining how the way Rap-acts attained status –either by

conforming or differentiating– moderates the relation between (non-)conforming behaviour

and song performance. Hypothesized is that acts that attained status through (non-)conformity

are more likely to perform better when (non-)conforming in the future. For all songs that

appeared on Billboard’s ‘Top Rap Song’ charts from 2006 up to and including 2016 it is calculated how much they differ from their genre’s standards based on their musicological features. A distinction is made between low-status, status through conforming, and

high-status through differentiating acts based on whether they were nominated for a Grammy

Award for Best Rap Album and whether the nominated act was conforming or differentiating

at that moment. The results indicated that acts that attained a high level of status through

conformity are more likely to perform better when conforming in the future. For the other

groups no significant results were found. Furthermore, the practical and theoretical

implications of the findings are discussed. Lastly, the limitations and suggestions for further

(5)

5

1. Introduction

1.1 Background

In 1973 DJ Kool Herc organized parties in the Bronx, New York and played music by doing

something that has never been done before. By using two turntables he managed to play the

beat break sections of multiple Funk, Soul, and Disco records via a technique called

‘looping’: One turntable plays the beat break section and when it ended the other turntable would start playing that same section creating a loop that could go on as long as DJ Kool Herc

wanted to. Herc invited Coke La Rock to do what we now refer to as ‘rap’ over the beat to

keep the people partying. At this moment the history of music was made as Rap came into

existence.

Rap got picked up by more acts (i.e. artists and groups) and it eventually lead

Sugarhill Gang to release the very first commercial Rap song ‘Rapper’s Delight’ in 1979

(Lena & Pachucki, 2013). Less than a decade later the album ‘Escape’ by Hip-Hop group Whodini would be the first Rap album to ever reach the top 40 in the Billboard Top Pop

charts. From this moment till the early 1990s is considered to be the ‘Golden Era of Rap’ (Caramanica, 2005). During this time, Rap dominated the pop charts and enjoyed its artistic

legitimation through acts like LL Cool J, Run-D.M.C., A Tribe Called Quest, N.W.A., and 2

Live Crew (Lena & Pachucki, 2013). The boundaries of the music genre called ‘Rap’ were established (Lena & Peterson, 2008) and in the year 1996 the National Academy of Recording

Arts and Sciences, known for organizing the Grammy Awards, created the category ‘Grammy Award for Best Rap Album’.

However, the boundaries of a genre are not fixed (Lamont & Mólnar, 2002) and

audiences will debate on whether particular songs or acts belong to a specific music genre

(Walser, 1993). This is also the case for Rap, as for the last decade there has been a lot of

(6)

6

genre when the audience perceives the act as conforming to the genre’s standard, which is set

by a set of aesthetic norms defining the genre. Acts can deviate from this standard and push

their genre’s boundaries. Although it is not expected that deviation will lead to success in general, it does increase the likelihood of exceptional success (Hsu, Negro, & Perretti, 2012).

Rap acts like De La Soul, Kanye West, and Childish Gambino are some of the acts

that have been expanding the boundaries of their genre and gained status and success by doing

so. However, not all the acts who attained status by deviating from the standards reach

success later in their careers, even though some of them started conforming to the new set of

rules within their genre while others kept on differentiating and pushing the boundaries even

further. Besides these acts, there are also artists and groups who attained status while they

were conforming to the rules of their genre (e.g., Common, Lupe Fiasco, and J. Cole).

However, just as with the differentiating acts, these conforming acts do not necessarily reach

further success by sticking to conforming.

Each act has a personal optimal point in balancing conforming to versus

differentiating from their genre’s codes. Reaching this optimal point will result in multiple benefits, such as earning more positive evaluations, enjoying lower competitive intensity, and

achieving higher performance (Deephouse, 1999).

Past research found that this optimal point differs per level of status (Phillips &

Zuckerman, 2001). Middle-status actors tend to conform, as they fear that differentiating from

their category’s standards will lead negative evaluations and a loss of status (Philips & Zuckerman, 2001). The high- and low-status actors do not feel constrained by evaluations of

others and thus experience a sense of freedom to differentiate from the rest (Duguid &

(7)

7

1.2 Problem Definition and Research Objective

In the previous studies (Duguid & Goncalo, 2015; Durand & Kremp, 2016; Phillips &

Zuckerman, 2001; Washington & Zajac, 2005) that investigated the effect of status on an

actor’s optimum in terms of (non-)conformity, the way in which these actors attained their status –either by conforming or differentiating– has not been investigated. In order to fill in

this gap in the literature, I would like to research, besides how the status-level of actors

impacts their optimal point of differentiation compared to the other actors in the same

category, whether the way in which an actor attained status –either by conforming to or

differentiating from the category’s codes– influences that actor’s necessity to conform or differentiate in the future. An attempt to put this ambition in the form of a question is

presented next: “How does the way in which actors attain status impact their future optimum

in terms of conforming to versus differentiating from the rules within their category?”

By finding an answer to this question I aim to fill in the aforementioned gap in the

literature, and to provide music artists and record labels with some guidance that they could

use in future decisions on whether they (or their artists) should conform to or differentiate

from the rules of the genres they operate in.

I expect that acts that attained status through conformity are more likely to perform

better when conforming in the future while acts that attained status through differentiating

from the standards will become more successful when differentiating in the future.

In order to test these predictions I will work with the database of AcousticBrainz.org

to collect musicological data on recordings (i.e. songs). This allows me to examine the

position a song takes within its genre and determine whether it is conforming to or

differentiating from that genre’s standards. Additionally, nominations for Grammy Awards

(8)

8

moment. Furthermore, I will observe how successful these acts are in terms of sales, streams,

and radio play by working with the database of Billboard.

Combining all these data I will be able to determine whether the acts’ songs were

conforming or differentiating at the time they were nominated for a Grammy, whether the

following works of these acts were conforming or differentiating from the new established

standards, and whether their (non-)conforming behaviour led to more or less success in the

future.

1.3 Structure

In the following section I will introduce the theoretical background on concepts such as,

genres and codes within genres, the trade-off between conforming and differentiating,

status-levels, and how these concepts are related to each other. Furthermore, I will present the

hypotheses of this research. After that, I will discuss the approach and methods of this

research in the research method section. Moreover, I will provide the found results. Finally,

this paper will come to a conclusion by discussing the implications, limitations, and areas for

(9)

9

2. Literature review

2.1 Genres and codes

Genres (i.e. categories) are established within networks through the collective understanding

of the processes and standardized classifications (i.e. codes) which make it possible for

participants and audiences to perceive similarities and distinguish differences between

products (Becker, 1982; Hsu & Hannan, 2005). Through applying these codes one is able to

determine whether a product shares similarities with products within its genre and whether it

is different from products from other genres, this enables consumers to value the product

accordingly (Aucouturier & Pachet, 2003; Hsu & Hannan, 2005). Applying the concept of

genre to the creative industries, it encompasses a system of orientations, conventions, and

expectations that not only links producers (i.e. artists) and consumers to each other; it

connects artists, peers, critics, fans, and other stakeholders by defining what they identify as a

distinctive sort of creative experience (Anand & Croidieu, 2015; Lena & Peterson, 2008). In

the creative industries, a substantial amount of time is devoted to distinguishing what is and is

not art, who is and is not an artist, and what kind of art is and is not theirs (i.e. part of their

genre) (Anand & Croidieu, 2015; Becker, 1982). By doing so, one can distinguish, for

example, Kitsch from high art, Quentin Tarantino’s movies from Steven Spielberg’s, but also,

Chicago Blues from Electric Blues. These distinctions are often based on socially accepted

rules (Fabbri, 1982), however, scientists and the creative industries’ commercial side tend to

make distinctions based on the intrinsic attributes of the works (Aucouturier & Pachet, 2003;

Lena & Peterson, 2008).

In the papers of Aucouturier and Pachet (2003), and Askin and Mauskapf (2017)

music genres were being dissected based on the features of musical works, such as the key,

tempo, and instrumentation. This way, Askin and Mauskapf (2017) were able to quantify the

(10)

10

genres (i.e. how different these songs are from the genre’s standards). These codes can be

expressed through, for example, the amount of ‘danceability’ of a song, around which tempo a

song should be played, or through the amount of ‘speechiness’ that is present throughout a

track (Askin & Mauskapf, 2017).

2.2 Conformity, differentiation, and trade-off propositions

2.2.1 Organizations should conform

An actor in a category is said to be conforming when this actor behaves in accordance to the

norms of the category it belongs to (Hollander, 1958). Much prior research argues that

conforming to established codes within a category is important as it leads to a higher

perceived legitimacy, better external evaluations, and better performance (Chen & Hambrick,

1995; Durand, Rao, & Monin, 2007; Kennedy, 2008; Pólos, Hannan, & Carroll, 2002). For

example, Durand et al. (2007) found that French chefs that conform to the codes of the

nouvelle cuisine enjoy significantly more positive evaluations than non-conforming chefs.

Being non-conforming, and thus not meeting the institutionalized expectations, would lead to

lower legitimacy (DiMaggio & Powell, 1983), which in turn leads to diminishing abilities to

acquire resources as the actors would be penalized or systematically ignored by intermediaries

in the industry (Hsu, Hannan, & Koçak, 2009; Zuckerman, 1999). Ultimately, this would

result in higher fail rates (Hsu et al., 2009; Kovács & Johnson, 2014; Zimmerman, 1999).

2.2.2 Organizations should differentiate

However, there has also been much research conducted on the positive impact that

differentiating from the consensus shared by the actors within a category has on the

performance of organizations (Deephouse, 1999; Durand & Paolella, 2013; Paolella &

Durand, 2016). Differentiating is reported to lead to lower competitive intensity, which

(11)

Ranger-11

Moore, & Banaszak-Holl, 1990). Furthermore, positive correlations between client

evaluations, performance, and differentiating organizations were reported in previous studies

(Deephouse, 1999; Durand & Paolella, 2013; Paolella & Durand, 2016). Moreover, Paolella

and Durand (2016) found that law firms that span multiple categories are more likely to obtain

positive assessments as these firms are more likely to live up to the audiences’ expectations.

Category-spanning firms are regarded as being more competent and having a clearer identity

than firms that focus on a single category (Paolella & Durand, 2016). Subsequently, these

positive assessments significantly mediate the relation between category spanning and

performance, and this relation is even stronger when the combination of different categories is

perceived as making more sense than other combinations (Paolella & Durand, 2016; Phillips,

Turco, & Zuckerman, 2013). For example, a combination of the music genres Country and

Rock would make more sense as it occurs more often than a combination of Rap and Classical

music, and is thus more likely to be received with positive criticism and higher sales. Hsu et

al. (2012), however, would argue that this uncommon combination of music genres is more

likely to lead to an exceptional success. For example, as seen in ‘I Can’, a Rap-song by Nas

which heavily samples Ludwig van Beethoven’s ‘Für Elise’. By reaching the twelfth position on Billboard’s Hot 100 chart it is Nas’s highest charting single thus far (Billboard, 2018).

2.2.3 Organizations should balance conformity and differentiation

Brewer (1991) was the first to introduce the theory of ‘optimal distinctiveness’, she argued that individuals pursue an optimal balance between being part of and being distinct from

social groups and situations. Deephouse (1999) introduced a similar kind of concept but

applied it to organizations. He argued that firms are pressured to conform to but also

differentiate from the standards, so these firms should seek an optimal balance between

(12)

12

they should “achieve maximum performance at the level of strategic similarity where the gains from reduced competition are equal to the costs of legitimacy challenges” (Deephouse, 1999, p. 154). This optimal balance concerning conformity and differentiation is context

specific and changes over time (Leonardelli, Pickett, & Brewer, 2010). Askin and Mauskapf

(2017) applied Deephouse (1999) his concept of optimal differentiation to music and

hypothesized that songs are able to manage a similarity-differentiation trade-off. Askin and

Mauskapf (2017) argued that songs’ position in the genre predicts their success in terms of

sales. They found that songs which sound too much alike previously released songs are less

likely to perform well, while songs that accomplish optimal differentiation are more likely to

succeed (Askin & Mauskapf, 2017).

2.3 Status and conformity

2.3.1 Attainment of status

According to Podolny (1993) the status of a producer in a market is defined as the perceived

quality of that producer’s products compared to the perceived quality of the competitors’ products. Status is derived from subjective evaluations from the audience on the actor’s actions within the category’s task structure and its expressive structure (Ridgeway, 1978). This evaluating audience consists of experts, peers, and/or consumers (Anderson & Kilduff,

2009; Duguid & Goncalo, 2015; Podolny, 1993; Yogev, 2010). Examples of indicators of

status are Michelin Stars rewarded to restaurants and chefs (Durand et al., 2007), Oscars given

to filmmakers, and Grammy Awards won by music acts (Anand & Watson, 2004).

Producers are argued to be able to attain status by conforming (Hollander, 1958, 1960)

or by non-conforming (Ridgeway, 1978, 1981; Wahrman & Pugh, 1972). Hollander (1958,

1960) stated that an actor attains and maintains increased status by conforming to the

expectations of that actor’s category, while differentiating from the category’s expectations will lead to a loss of status. Hollander (1958, 1960) argued that this attained status is formed

(13)

13

by the positive evaluations the rest of the group applies to the individual actor. The amount of

status an actor enjoys is reflected in the degree to which this actor is able to deviate from the

category codes without any punishment (Hollander, 1960).

Wahrman and Pugh (1972) replicated the study of Hollander (1958) and found

different results: Where Hollander (1958, 1960) stated that deviating from the category’s standards had negative effects on the level of status attained, Wahrman and Pugh (1972)

found that non-conformity actually leads to a positive effect on the level of status earned. This

positive effect is due to the fact that non-conforming actions attract the category’s attention

and thus facilitate awareness of that actor’s actions and contributed competence (Wahrman & Pugh, 1972; Ridgeway, 1981).

2.3.2 Levels of status and conformity

Some studies made a distinction in levels of status; namely low-status, middle-status, and

high-status, which subsequently were related to the conformity equilibrium (Duguid &

Goncalo, 2015; Durand & Kremp, 2016; Phillips & Zuckerman, 2001; Washington & Zajac,

2005). It is argued that producers with middle-status are aware and concerned about their

position in the status hierarchy which will deter them from differentiating as they fear

negative evaluations and eventually lower status and bad future performance (Galinsky,

Magee, Gruenfeld, Whitson, & Liljenquist, 2008; Phillips & Zuckerman, 2001).

High-status producers, however, are less concerned and thus less constrained by the

evaluations of others, so they enjoy more freedom to differentiate and come up with creative

ideas in the future (Duguid & Goncalo, 2015). Washington and Zajac (2005) hold a different

view as they expect that high-status producers perceive higher expectations from their

(14)

14

For low-status producers this feeling of freedom to not conform to the group’s

expectations coincides with the feeling enjoyed by their high-status peers; they do not feel the

risks that comes with deviating from the standards as they are still at the bottom of the status

hierarchy and therefore have less to lose (Duguid & Goncalo, 2015).

2.4 Hypotheses

Previous studies found a relation between the status-levels of actors and their optimal future

behaviour with regard to (non-)conformity. However, no study has researched whether the

way in which actors attain their status affects their ideal future behaviour in terms of

conforming versus differentiating. Where past literature (Duguid & Goncalo, 2015; Durand &

Kremp, 2016) stated that high-status actors should rather differentiate from than conform to

the standards, I expect that these actors’ optimal behaviour is affected by the nature of their

status. If an actor attains status through, for example, critic reviews, it attracts attention to the

actor’s work and leaves the consumer with impressions and future expectations (Eliashberg & Shugan, 1997). If the consumer’s impression is that the actor deviated from the standards,

then it could be the case that further deviation will lead to the actor meeting its audience’s

expectations, which in turn results more favourable evaluations (Durand et al., 2007). In other

words, I expect that the actors’ past behaviour that gained them critical acclaim influences

their optimal point of (non-)conformity in the future. Therefore, it is necessary to test the

following hypotheses:

H1: Acts that attained a high level of status through conformity are more likely to perform better when conforming in the future.

H2: Acts that attained a high level of status through differentiating are more likely to perform better when differentiating in the future.

(15)

15

3. Research Method

This study is based on quantitative database research. The data is extracted from multiple

databases. The used databases, research design, chosen samples, measures, and variables will

be elaborated on in this chapter.

3.1 Sample and Data Collection

This study will be performed within the music industry, as it is believed to represent the ideal

context to test the relative typicality of products (Askin & Mauskapf, 2017). While songs can

differ a lot from each other, they still adhere to the same standards in terms of rhythm,

harmony, and melody (Askin & Mauskapf, 2017). In this paper it will be investigated how

music acts’ status moderates the relation between their behaviour and performance. I will focus on the relative younger music genre called Rap, as it is still evolving and as it

encounters a lot of discussion on what its standards and boundaries are (Penrose, 2017;

Turner, 2017; Watson, 2016). Due to data availability and a limited amount of time to conduct

this research, I will focus on acts that appeared on Billboard’s ‘Hot Rap Songs’ chart from

2006 up to and including 2016. By using this chart, it is ensured that the acts in this study are

commensurable in terms of behaviour and performance, and were active during the period

examined.

Information of the acts’ performance in terms of sales, streams, and radio play is retrieved from the Billboard database. The charts of Billboard are seen as the most reliable

performance measurers in the music industry (Askin & Mauskapf, 2017; Rossman, 2012).

Another database that will be used is the one of AcousticBrainz.org. AcousticBrainz is

a platform where users utilize procedures and algorithms from the Music Technology Group

at Pompeu Fabra University (Bogdanov et al., 2013) in order to document the musicological

(16)

16

AcousticBrainz, the musicological features such as, the key, tempo and danceability of the

songs that entered the Billboard chart can be found. This information will be used to

determine the standards of the genre per feature per year. That wayit is possible to assess to

what extent each song is (non-)conforming to the standards.

In addition, the database of The National Academy of Recording Arts & Sciences will

be used in order to retrieve information on whether the acts obtained any nominations for

‘Grammy Awards for Best Rap Album of The Year’, indicating critical acclaim and a high status-level within the music industry (Anand & Watson, 2004; Caves, 2000).

For these Grammy nominees, it will be investigated whether they attained their

high-level status through conformity or differentiation. The next step is to investigate the amount of

()conformity and the performance of their following releases; both charting and

non-charting songs. Furthermore, a conclusion will be drawn on the relation between the

independent behaviour variable and the dependent performance variable as a function of the

moderating status variable in order to establish the artists’ optimal point in terms of

(non-)conformity. This way it will be possible to conclude whether acts should stick to conforming

or differentiating in the future or whether they should switch up their behaviour.

3.2 Measurement of Variables

3.2.1 Dependent Variable

Performance is measured as the weighted average of sales, airplay, and streams provided by

the weekly Billboard ‘Top Rap Songs’ chart. As done by Berger and Le Mens (2009), longevity (i.e. weeks on chart) will be used as the dependent variable to measure the

(17)

17

3.2.2 Independent Variables

By using the AcousticBrainz API it is possible to collect information on the musicological

features of the songs that entered the chart. These features are attributes like the key, scale,

and chord the song is played in. These attributes are denoted in words, so it is necessary to

recode them into dummy variables in order to make them measureable (Askin & Mauskapf,

2017). The songs’ length, ‘danceability’, beat count, and beats per minute (BPM) are also

observed. All the attributes taken into account in this research are presented and briefly

explained in Table 1.

It is necessary to determine how much each song i differs from the songs that appeared

on the chart in the year before song i’s debut. To do this, each song’s attributes were

translated into numbers and then normalized (i.e. dividing the values by the highest observed

value for each attribute) so they are valued strictly between 0 and 1. Subsequently, the

averages per attribute per year were computed to determine each year’s standards. After that,

the squared difference between the attributes of each song i and the attributes of the average

song released in the year prior to song i’s debut was calculated. The last step contained

finding the square root of the sum of these squared differences, resulting in the Euclidean

distance between each song i and the average song of the year prior to song i’s debut. By

calculating this distance it is possible to determine how much a song is conforming to or

differentiating from the genre’s standards. Where Askin and Mauskapf (2017) measured the difference by calculating the cosine similarity of the songs, I calculated the Euclidean distance

as it takes into account the actual distance between vectors where the cosine similarity only

looks at the angle between vectors (Emmery, 2017). The Euclidean distance got

mean-centered so that a value of 0 would mean that an artist’s behaviour in terms of conformity is

average. As an inverted U-shaped relationship is expected, a new variable is computed by

(18)

18

3.2.3 Moderating variable

The moderating variable, status, is based on an act’s (non-)conforming behaviour and the critical acclaim the act has attained. In prior research, distinctions are made between high,

middle, and low status-levels (Duguid & Goncalo, 2015; Phillips & Zuckerman, 2001). In this

research however, a different distinction will be made between the acts: The first group (1)

contains artists that attained high status by conforming to the standards of the genre at that

time. The second group encompasses acts that also attained high-status, however, through

differentiating from the genre’s standards (2). The baseline group (0) consists of acts that did

(19)

19

performance of high-status acts, and as it will save time and effort not having to find out

whether acts are part of the low- or middle-status groups.

In this paper, nominations for Grammy Awards indicate high-level status as Grammy

Awards are the most well-established awards in the music industry signifying critical success

(Anand & Watson, 2004; Caves, 2000). If an act has never been nominated for this Grammy

Award, then it will be categorized in the low-status group. For the high-status-artists a

distinction must be made in terms of behaviour. To establish which distance corresponds to

which behaviour, it is necessary to determine the midpoint of these acts’ distances. The

median of all song’s distances was computed; 1.455198972. If the original distance of an act’s first charting song was shorter than the median distance, then this artist was coded as

conforming, while the original distance of differentiating acts was longer than the median

distance. It is expected that acts will perform better when (non-)conforming after they have

attained a high status-level for releasing a (non-)conforming product. In other words, acts are

expected to perform better when they stick to the behaviour that attained them critical

acclaim.

3.2.4 Control Variables

Multiple control variables are included to make the analysis less biased. The first variable that

controls for the advantages acts experience through prior visibility, popularity, and experience

takes into account how many songs an act had entering the chart before each new released

song. Similar to the research of Askin and Mauskapf (2017), this relative popularity an act

enjoys prior to a song’s release is ranked into four different levels: (1) for the first song that enters the chart, (2) for the second or third song that reaches the chart, (3) for the fourth

through tenth song on the chart, or (4) if the act already had over ten songs charting before.

Second, a dummy variable takes into account whether or not the charting song is part

(20)

20

year, as being nominated for a Grammy Award increases an artist’s visibility (Anand and

Watson 2004). The variable is coded (0) for non-nominees, and (1) for nominees.

Furthermore, a control variable accounting for the number of Grammy Award

nominations an act had earned before the moment its song charted also takes into account the

act’s increased visibility. This relative visibility is ranked into five different levels: (0) for acts that were never nominated for a Grammy Award before, (1) for acts with one prior

nomination, (2) for acts that were nominated two or three times before, (3) when the act is

nominated four or five times before, and (4) for artists that have been nominated over five

times before. Anand and Watson (2004) argued that winning a Grammy Award, but also

being nominated for one, positively impacts the sales of the act’s songs.

A dummy variable which is also related to an act’s prior experience, visibility, and

popularity is encompassed in the ‘multiple memberships’ variable. It accounts for acts that that released songs under different names or band formations (Askin & Mauskapf, 2017). For

instance, both Fergie and Will.i.am have songs that entered the charts as a solo artist;

however, these two artists became famous for being part of the music group called The Black

Eyed Peas. The variable is coded as (1) for acts like this and the count in amount of Grammy

Award nominations and previous charting songs is continued as these acts still enjoy their

prior experience, fans’ loyalty, and other potential benefits (Askin & Mauskapf, 2017). Another dummy variable takes into account the extra visibility and anticipation an act

enjoys when releasing a song with a feature of a top-tier artist (1). For example, Big Sean

released ‘My Last’ with a feature of Chris Brown, who by that time had multiple number one singles and was that year’s winner of the Grammy Award for Best R&B Album. The song was Big Sean’s first charting song and went straight to the number one spot.

The last dummy variable that is included takes into account whether an act released its

(21)

21

via a major label brings many advantages, as these labels have larger budgets for marketing

and production, stronger connections with bigger acts and radio stations, and more experience

in creating chart-friendly songs (Askin & Mauskapf, 2017; Rossman, 2012) making it easier

to enter the charts and reach the top. Table 2 presents the means, standard deviations, minima,

maxima and correlations for all the variables in the analysis.

4. Analysis

Because the dependent variable represents the amount of weeks a song is charting on the

Billboard chart and only can take non-negative integer values, a regression method suitable

for count data is necessary (Coxe, West, and Aiken, 2009). The Poisson model is chosen as it

is the standard solution and as it is said to perform better than other conventional model-based

methods in the literature (Cameron and Trivedi, 1998; Zhu, 2012). The equation used is:

ln(Y) = α + β1X1 + β2X22 + β3M + β4X1M + β5X22M + βZ + ε

Where Y is the amount of weeks that a song is charting on the Billboard chart for ‘Top Rap Songs’, α is the constant term, and β’s are the coefficients that will be estimated. X1 is the

mean-centered Euclidean distance between each song i and the average song of the year prior

to song i’s debut, and X22 represents the squared mean-centered Euclidean distance. M is the

moderating status variable, Z is a set of control variables, and ε is the standard error.

In total, seven Poisson regressions will be run with the amount of weeks on the chart

as the dependent variable. The first only including the control variables, the second adding the

status variable, the third adding the first independent variable, and the fourth includes the

other independent variable as well. For the fifth, sixth, and seventh model a sample is taken

including only the low-status group, high-conforming group, and high-differentiating group

(22)

(23)

23

5. Results

In order to test the two hypotheses seven Poisson regression analyses were conducted. Table 3

presents the results from these Poisson regression analyses. The beta-coefficients,

significance, and standard errors are provided for each model. Furthermore the number of

observations and the log-likelihood are presented as well. The latter is used to compare

models for the same data. When the log-likelihood of one model increases (i.e. moves towards

0) compared to the log-likelihood of another model with the same data, it suggests that this

model is more plausible than the other. Looking at the log-likelihood of the first four models,

it shows that the models are becoming a bit more plausible the more variables are added. For

the last three models, the log-likelihood enjoyed a much greater increment compared to the

first four models.

It was hypothesized that an inverted-U shape relation would be found for the two

high-status groups. Lind and Mehlum (2010) proposed a three-step procedure to determine whether

a relationship is actually quadratic and whether its shape is inverted or not. First of all, the

beta of the quadratic function (β2) must be significant and of the expected sign. A negative β2,

in combination with a positive β1, corresponds with an inverted-U shape relationship, and a

positive β2, combined with a negative β1, relates to a regular U-shaped one. Second, the slope

must be significantly different from zero at the low- and high end of the X-range. Third, the

turning point of the function should be within the data range. Looking at the coefficients of

model 6, the relation seemed to be U-shaped rather than inverted-U shaped. This means an

optimal point in terms of conforming to versus differentiating from the standards cannot be

found, however the worst point could be determined though. Besides β2 of model 7 having the

opposite sign, the coefficient was not significant. In Figure 1 the plots of the graphs for each

(24)

(25)

25

Figure 1. Plots of Poisson regressions per status group

(model 5), the solid green line corresponds with the model of the high-conforming group

(model 6), and the dotted black line resembles the plot for the group that attained high-status

through differentiating (model 7). The circles represent the turning points of the models. As

the data ranges from -0.631 to 0.448 the plot is cut off around there as these points represent

the boundaries of the Rap genre. In other words, songs that are more conforming or more

differentiating than this range allows for do not occur in the observed data. This could imply

that songs that are more differentiating are not recognized as Rap anymore but rather a

different genre, and thus tend to be overlooked by consumers, critics, and peers (Kennedy,

2008).

Only the model corresponding with the group that attained high-status through

conforming to the standards (model 6) is significant. To determine whether this group’s model

significantly differs from the others’ it is necessary to determine whether the coefficients of its independent variables lie outside the confidence interval of the coefficients of the other

(26)

26

models. In table 5 each group’s 95% Wald confidence interval can be found along with each

group’s betas. The low-status group’s betas can be found within the confidence interval of the high-differentiating group, and vice versa. This implies that these two groups do not

significantly differ. The coefficients of the high-conforming group (model 6) lie outside the

other group’s confidence intervals, so it can be concluded that this model is significantly different than the other two groups.

Knowing that model 6 is meaningful as its Omnibus test was significant with p < .000,

and is significantly different than the other groups’ models we can interpret the shape of its curve. The little green circle in figure 1 shows that the least optimal point is situated within

the differentiating area of the graph. As the line lies higher on the left side of the mean, we

can say that acts from this high-conforming group will perform better when they conform

more to the set standards than the average artist. The more these acts conform the better they

will perform, up until a certain point at least. Their performance decreases as they start

differentiating, however, it increases somewhat again after a certain point. Besides this

group’s behaviour in terms of (non-)conforming, some other factors significantly influence their performance: Being currently nominated for a Grammy (β = 0.175; p < 0.01) having one

to three previous Grammy nominations (β = 0.306, β = 0.303; p < 0.01), and having more than

(27)

27

compared to having no Grammy nominations and only one charting song. Unexpectedly,

having over five Grammy nominations has a significant negative effect (β = 0.201; p < 0.01)

on the conforming high-status group’s performance.

Although the expected inverted-U shaped relationship was not found, the results did

provide support for the first hypothesis, which claimed that artists who attained a high level of

status through conformity are more likely to perform better when conforming in the future.

The other hypothesis, which claimed that acts that attained a high level of status through

differentiating are more likely to perform better when differentiating in the future, cannot be

accepted as the behaviour coefficients of this group’s model (model 7) were insignificant and

as this model was not significantly different than the model of the low-status group (model 5).

6. Discussion

In the next section, the aims of the current study will be repeated, and the findings will be

discussed. Furthermore, both the theoretical and managerial implications will be elaborated

on. Additionally, the limitations of the current study will be clarified, and suggestions for

further research will be provided.

Given the role of status in actor’s behaviour, performance, and longevity (Galinsky et al., 2008; Phillips & Zuckerman, 2001), there is both managerial and academic relevance in

understanding what strategic behaviour might be preferred in order to stimulate one’s performance. Where it has been stated that high-status actors are best of by differentiating

from the rest of their market category (Duguid & Goncalo, 2015; Durand & Kremp, 2016),

there has been no research into the nature of an actor’s high-status attainment, and how this

impacts the relation between (non-)conforming behaviour and performance. Therefore, the

aim of the current study was to fill this gap in the theory by analysing the role of a high

(28)

status-28

level. Besides the gap in the literature, a managerial gap was aimed to be filled. Aimed for

was that, with the results of this research, actors would be provided with some sort of

guidance to help them with future strategic decisions in terms of (non-)conforming behaviour.

By collecting data from Billboard and AcousticBrainz, it was possible to find support

for hypothesis 1 (actors who attained a high level of status through conformity are more likely

to perform better when conforming in the future). While the consensus in the literature is that

high-status actors should differentiate in order to perform better (Duguid & Goncalo, 2015;

Durand & Kremp, 2016), the findings of the current study differ: As a distinction was made

between acts that attained high-status through conformity or differentiation, it was found that

it depends on the high-status actor’s nature whether conforming or differentiating results in

better performance. The findings that support hypothesis 1 correspond with the findings of

Washington and Zajac (2005) who claimed that high-status producers perceive higher

expectations from their audience and thus, in order to perform well, will choose to conform to

those expectations in the future. As Phillips and Zuckerman (2001) found that there is an

inverted-U shaped relationship between status and (non-)conformity, it was expected to find

an inverted-U shaped relationship between behaviour and performance in the current study as

well. However, it appeared to be non-existent as a U-shaped relationship was found.

Additionally, no evidence was found to support the expectations that actors who attained a

high status-level through differentiating are more likely to perform better when differentiating

in the future. So, this current research showed that not only an actor’s status-level impacts the

relation between (non-)conformity and performance, but that the nature of the actor’s status

attainment plays a significant role as well.

However, there are several limitations that need to be addressed. The first, being the

nature of the sample: By only focussing on one genre that is relatively young, the findings

(29)

low-29

level information on the song’s musicological features was used, so high-level information (e.g. the timbre, mood, and model of a recording) was ignored. Other important features for

the Rap genre, such as the rhyme scheme, delivery, and topic, were not taken into account

either. In this study, I did not control for whether an artist performed during the Grammy

Award ceremonies, which was found to increase sales (Anand & Watson, 2004). Furthermore,

only the amount of weeks a song spent on the chart was observed to measure its performance.

The position a song held from week to week was not taken into consideration to save time and

effort. In addition, the Poisson regression models used in this study were overdispersed.

Overdispersion leads to optimism and thus an inflated type I error rate. A quasi-Poisson

regression would have been a better option; however, the statistical software used for this

study is not able to conduct this type of analysis. Finally, the second step of the three-step

procedure of Lind and Mehlum (2010) could not be carried out as intended, so whether the

slopes at both ends of the data range were significantly different from zero remains unknown.

In future research, one should observe multiple genres (i.e. categories) within an

industry, take into account more detailed information, and make sure to be able to perform the

right analyses. Besides that, one could research how status-levels play a role in the relation

between spanning multiple genres and performance, while also taking into account which

genres are popular or up-and-coming at the time. Furthermore, it could be interesting to see

how the expectations of critics and consumers differ for each status group, and how these

expectations in turn influence an actor’s behaviour. This could clarify the reasons why some

actors choose to differentiate from others or not, and whether these decisions lead to a better

performance. In other words; do actors conform to the rules of their category or do they only

conform to the expectations of the audience regardless of their genre’s norm? What do actors

perceive as their role, and how does their status-level impact the relation between their

(30)

30

In conclusion, this research has contributed to the literature by assessing the

moderating effect of the way in which actors attained a high status-level in the relation

between (non-)conforming behaviour and performance. The presented findings may provide

actors in the music industry with some guidance to help them decide on their future strategies.

Additionally, the findings may initiate new ideas for future research on the relation between

(31)

31

References

- Anand, N., & Croidieu, G. (2015). Niches, Genres, and Classifications in the Creative

Industries. In The Oxford Handbook of Creative Industries (pp. 327-347). Oxford University

Press Oxford.

- Anand, N., & Watson, M. R. (2004). Tournament rituals in the evolution of fields: The case

of the Grammy Awards. Academy of Management journal, 47(1), 59-80.

- Anderson, C., & Kilduff, G. J. (2009). The pursuit of status in social groups. Current

Directions in Psychological Science, 18(5), 295-298.

- Askin, N., & Mauskapf, M. (2017). What makes popular culture popular? product features

and optimal differentiation in music. American Sociological Review, 82(5), 910-944.

- Aucouturier, J. J., & Pachet, F. (2003). Representing musical genre: A state of the art.

Journal of New Music Research, 32(1), 83-93.

- Becker, H. S. (1982). Art worlds. University of California Press.

- Berger, J., & Le Mens, G. (2009). How adoption speed affects the abandonment of cultural

tastes. Proceedings of the National Academy of Sciences, 106(20), 8146-8150.

- Bogdanov, D., Wack, N., Gómez Gutiérrez, E., Gulati, S., Herrera Boyer, P., Mayor, O., ...

Serra, X. (2013). Essentia: An audio analysis library for music information retrieval. In de

Souza Britto A, Gouyon F, Dixon S (Eds.), Proceedings of the 14th Conference of the

International Society for Music Information Retrieval (pp. 493-498) Curitiba, Brazil:

International Society for Music Information Retrieval.

- Billboard (2018, June 20). Nas I Can Chart History. Retrieved from:

https://www.billboard.com/music/Nas/chart-history/hot-100/song/429740

- Brewer, M. B. (1991). The social self: On being the same and different at the same time.

(32)

32

- Cameron, A. C., & Trivedi, P. K. (2013). Regression analysis of count data (Vol. 53).

Cambridge university press.

- Caramanica, J. (2005, June 26). Hip-Hop’s Raiders of the Lost Archives. The New York

Times. Retrieved from:

http://www.nytimes.com/2005/06/26/arts/music/hiphops-raiders-of-the-lost-archives.html

- Carroll, G. R., & Hannan, M. T. (1989). Density delay in the evolution of organizational

populations: A model and five empirical tests. Administrative Science Quarterly, 411-430.

- Caves, R. E. (2000). Creative industries: Contracts between art and commerce (No. 20).

Harvard University Press.

- Chen, M. J., & Hambrick, D. C. (1995). Speed, stealth, and selective attack: How small

firms differ from large firms in competitive behavior. Academy of management journal, 38(2),

453-482.

- Cordon, K. (2016, December 1). Examining the 2016 XXL Freshman Class of Hip Hop.

Retrieved from:

https://studybreaks.com/culture/examining-the-2016-xxl-freshman-class-of-hip-hop/

- Coxe, S., West, S. G., & Aiken, L. S. (2009). The analysis of count data: A gentle

introduction to Poisson regression and its alternatives. Journal of personality assessment,

91(2), 121-136.

- Deephouse, D. L. (1999). To be different, or to be the same? It’s a question (and theory) of

strategic balance. Strategic management journal, 20(2), 147-166.

- DiMaggio, P., & Powell, W. W. (1983). The iron cage revisited: Collective rationality and

institutional isomorphism in organizational fields. American Sociological Review, 48(2),

147-160.

- Duguid, M. M., & Goncalo, J. A. (2015). Squeezed in the middle: the middle status trade

(33)

33

- Durand, R., & Kremp, P. A. (2016). Classical deviation: Organizational and individual status

as antecedents of conformity. Academy of Management Journal, 59(1), 65-89.

- Durand, R., & Paolella, L. (2013). Category stretching: Reorienting research on categories in

strategy, entrepreneurship, and organization theory. Journal of Management Studies, 50(6),

1100-1123.

- Durand, R., Rao, H., & Monin, P. (2007). Code and conduct in French cuisine: Impact of

code changes on external evaluations. Strategic Management Journal, 28(5), 455-472.

- Eliashberg, J., & Shugan, S. M. (1997). Film critics: Influencers or predictors?. The Journal

of Marketing, 68-78.

- Emmery, C. (2017, March 25). Euclidean vs. Cosine Distance. Retrieved from:

https://cmry.github.io/notes/euclidean-v-cosine

- Fabbri, F. (1982). A theory of musical genres: two applications. Popular music perspectives,

1, 52-81.

- Galinsky, A. D., Magee, J. C., Gruenfeld, D. H., Whitson, J. A., & Liljenquist, K. A. (2008).

Power reduces the press of the situation: implications for creativity, conformity, and

dissonance. Journal of personality and social psychology, 95(6), 1450.

- Hannan, M. T., Ranger-Moore, J., & Banaszak-Holl, J. (1990). Competition and the

evolution of organizational size distributions. Organizational evolution: New directions,

246-268.

- Hollander, E. P. (1958). Conformity, status, and idiosyncrasy credit. Psychological Review,

65(2), 117-127.

- Hollander, E. P. (1960). Competence and conformity in the acceptance of influence. The

Journal of Abnormal and Social Psychology, 61(3), 365-369.

- Hsu, G., & Hannan, M. T. (2005). Identities, genres, and organizational forms. Organization

(34)

34

- Hsu, G., Hannan, M. T., & Koçak, Ö. (2009). Multiple category memberships in markets:

An integrative theory and two empirical tests. American Sociological Review, 74(1), 150-169.

- Hsu, G., Negro, G., & Perretti, F. (2012). Hybrids in Hollywood: a study of the production

and performance of genre-spanning films. Industrial and Corporate Change, 21(6),

1427-1450.

- Kennedy, M. T. (2008). Getting counted: Markets, media, and reality. American sociological

review, 73(2), 270-295.

- Kovács, B., & Johnson, R. (2014). Contrasting alternative explanations for the consequences

of category spanning: A study of restaurant reviews and menus in San Francisco. Strategic

Organization, 12(1), 7-37.

- Lamont, M., & Molnár, V. (2002). The study of boundaries in the social sciences. Annual

review of sociology, 28(1), 167-195.

- Lena, J. C., & Pachucki, M. C. (2013). The sincerest form of flattery: Innovation, repetition,

and status in an art movement. Poetics, 41(3), 236-264.

- Lena, J. C., & Peterson, R. A. (2008). Classification as culture: Types and trajectories of

music genres. American sociological review, 73(5), 697-718.

- Leonardelli, G. J., Pickett, C. L., & Brewer, M. B. (2010). Optimal distinctiveness theory: A

framework for social identity, social cognition, and intergroup relations. Advances in

experimental social psychology, 43, 63-113.

- Lind, J. T., & Mehlum, H. (2010). With or without U? the Appropriate Test for a U‐Shaped

Relationship. Oxford bulletin of economics and statistics, 72(1), 109-118.

- Paolella, L., & Durand, R. (2016). Category spanning, evaluation, and performance: Revised

theory and test on the corporate law market. Academy of Management Journal, 59(1),

(35)

35

- Penrose, N. (2017, May 15). Waka Flocka Flame Says Lil Uzi Vert is ‘Rock Not Hip-Hop’.

Retrieved from:

https://www.billboard.com/articles/columns/hip-hop/7793157/waka-flocka-flame-lil-uzi-vert-rock-not-hip-hop

- Phillips, D. J., Turco, C. J., & Zuckerman, E. W. (2013). Betrayal as market barrier:

Identity-based limits to diversification among high-status corporate law firms. American

Journal of Sociology, 118(4), 1023-1054.

- Phillips, D. J., & Zuckerman, E. W. (2001). Middle-status conformity: Theoretical

restatement and empirical demonstration in two markets. American Journal of Sociology,

107(2), 379-429.

- Podolny, J. M. (1993). A status-based model of market competition. American journal of

sociology, 98(4), 829-872.

- Pólos, L., Hannan, M. T., & Carroll, G. R. (2002). Foundations of a theory of social forms.

Industrial and Corporate Change, 11(1), 85-115.

- Ridgeway, C. L. (1978). Conformity, group-oriented motivation, and status attainment in

small groups. Social psychology, 175-188.

- Ridgeway, C. L. (1981). Nonconformity, competence, and influence in groups: A test of two

theories. American Sociological Review, 333-347.

- Rossman, G. (2012). Climbing the charts: What radio airplay tells us about the diffusion of

innovation. Princeton University Press.

- Slavik, N. (2010, March 2). Grading XXL’s Top 10 Freshman List. Retrieved from:

https://djbooth.net/features/grading-xxls-freshman

- Turner, D. (2017, June 1). Look At Me!: The Noisy, Blown-Out SoundCloud Revolution

Redefining Rap. Retrieved from:

(36)

36

- Wahrman, R., & Pugh, M. D. (1972). Competence and conformity: Another look at

Hollander's study. Sociometry, 376-386.

- Walser, R. (1993). Running with the devil: Power, gender, and madness in heavy metal

music. Wesleyan University Press.

- Washington, M., & Zajac, E. J. (2005). Status evolution and competition: Theory and

evidence. Academy of Management Journal, 48(2), 282-296.

- Watson, D. A. (2016, January 18). The Evolution Of Hip-Hop. Retrieved from:

https://www.theodysseyonline.com/the-evolution-of-hip-hop

- Yogev, T. (2009). The social construction of quality: status dynamics in the market for

contemporary art. Socio-Economic Review, 8(3), 511-536.

- Zhu, F. (2012). Modeling overdispersed or underdispersed count data with generalized

Poisson integer-valued GARCH models. Journal of Mathematical Analysis and Applications,

389(1), 58-71.

- Zuckerman, E. W. (1999). The categorical imperative: Securities analysts and the

To fit in, or to stand out? : the moderating role of status on the relation between (non-)conformity and rap song performance