The Effect of Executive Compensation
on Performance in the Creative Industries
26-08-2015
Malou Coopmans
MSc Business Administration
Entrepreneurship and Management in the Creative Industries 6032710
First Supervisor
E. (Erik) Dirksen MSc Second Supervisor
Statement of Originality
This document is written by Malou Coopmans who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
This study investigates the relation between executive compensation and firm performance in a creative and (semi-)public sector. From the executive compensation literature there is no consensus on whether pay-for-performance is effective. Studies have shown proof for a positive relation, however mostly in the US and with publicly listed firms. The setting in this research differs because it focuses on Europe, on a creative sector and on organizations with the government as main shareholder. For both the United Kingdom and the Netherlands, the national Public Service Broadcastings are investigated with regards to economic and artistic performance. The research hypothesizes a positive pay-for-performance according to the Efficient Contracting Theory, but also hypothesizes a stable economic performance according to the creative industries’ literature. The results show no relation between executive compensation and performance, neither economic nor artistic performance. The research investigates theoretical predictions using executive compensation data from 2006-2013 from the British and Dutch Public Service Broadcasting.
Table of Contents Statement of Originality 2 Abstract 3 1. Introduction 5 2. Literature review 11 2.1 Executive Compensation 12 2.2 Creative Industries 14
2.4 Public Service Broadcasting 19
2.5 Hypotheses 22
2.5.1 Economic Performance 22
2.5.2 Artistic performance 25
2.6 Conceptual Framework 26
3. Data and Methodology 27
3.1 Sample 27 3.2 Regression Analysis 30 3.3 Descriptive Statistics 33 3.4 Correlation Matrices 42 4. Results 45 4.1 NPO 45 4.2 BBC 47 5. Discussion 49 6. Conclusions 54 7. Reference list 58 Appendix 62
Exhibit 1: Awards won 62
1. Introduction
In 2007, the CEO (Chief Executive Officer) of UK firm Reckitt Benckiser received an executive compensation package of £ 31 million (Goergen & Renneboog, 2011). That same year, CEO of Dutch firm Shell earned £ 17.7 million (Goergen & Renneboog, 2011). These amounts of income are far higher than the amounts other employees in those companies earn. This raises questions like: are these amounts justifiable? Does the executive actually create so much value? While these types of publicly traded companies have been researched extensively, excessive compensation is spreading to companies in the much less researched sector, the (semi-)public sector.
In the Netherlands, the public pressured the Dutch government to act on excessive compensation in (semi-)public sector. The Dutch government had been pressured to constrain compensation laws, when the public discovered that part of their taxes is allocated to high earners in the Public Service Broadcasting (NRC, 2007). During the recent Financial Crisis, the public was offended by the government’s actions to cut budget for concerns like education and health care, when there are excessive compensation packages distributed in the (semi-) public sector. The Dutch government’s first reaction was in 2006 via a new policy (WOPT1), forcing government officials to report on any executive compensation higher than 130% of a minister’s salary. The Dutch government enforced the law to let the public criticize the excessive amounts and informally force the Public Service Broadcasting to lower excessive compensation. However, the public demanded more actions from the Dutch government, to which it responded by another policy change (WNT2). This law set a limit on compensation in the public and semi-public sector of 130% of the
1 Wet Openbaarmaking uit Publieke middelen gefinancierde Topinkomenst: every official employed by the government earning more than 130% of a minister’s salary needs to publicly disclose his or her income.
minister’s salary (called the “Balkenendenorm” at the time). Ultimately, the government decided that executive compensation in the public and semi-public sector should be limited to the minister’s compensation (Rijksoverheid, 2015a). A similar discussion went on in the United Kingdom, although no law as restricted as the Dutch law WNT was put in place.
Since the (semi-)public sector is rather large, this research focuses on the Public Service Broadcasting. The Public Service Broadcasting falls under the category creative industries, which combines two fields in which executive compensation is a (rather) new topic.
The Public Service Broadcasting has a merit good character. A merit good has the feature that the good should be available to everyone and serviced or financed by the government. The public benefits in its entirety from the fact that high quality broadcasting output is available (Hastings, 2004).
A Public Service Broadcasting can be funded via three options, pure public or pure commercial funding or a mix of both (European Broadcasting Union, 2015). Public funding may come from a license fee or directly from the state budget. For some countries, like the Nordic countries and the UK, pure public funding is sufficient. Still, most countries also rely on revenues coming from advertising.
Two nations are central in this research, namely the UK and the Netherlands. Both the UK and the Netherlands rely on different types of funding. The UK uses a license fee for funding the British Broadcasting Corporation, hereafter the BBC. The license fee has the advantage that it is independent of political interference and establishes a direct link between the public and the broadcaster. The Dutch Public
Service Broadcasting, the Nederlandse Publieke Omroep, hereafter the NPO, receives funding directly from the government. Direct funding has the advantage that no special collection or enforcement mechanisms are required. The license fee is perceived to be more stable and predictable than the government funding, since it is not subject to budget cuts (European Broadcasting Union, 2015). Previously the Dutch Public Service Broadcasting operated via a license fee, yet it changed to direct funding in 2000 (European Broadcasting Union, 2015). The change in funding led to high budget cuts initially, and also more recently in 2010 and 2012 (European Broadcasting Union, 2015).
Both nations use public funding in order to fund the Public Service Broadcasting and operate under control of the government.
The executive compensation in the Public Service Broadcasting is part of the literature on contracting theory. The Efficient Contracting Theory, which is more carefully explained in the literature review, claims that the current level of executive compensation indicates an equilibrium on the market for managerial talent and entails economic incentives to optimize firm value (Murphy, 2012). This theory comes from agency theory. The agency theory defines a gap between shareholders and managers, where managers control shareholders’ funds and need to be incentivized correctly to return wealth to the shareholders. The motivation for managers to optimize firm value can be put in the contract via stock options or other economic incentives that imply pay for performance. One problem with this theory is that the shareholder in this research does not want to optimize firm value for an increase of its own wealth. The government is not interested in a return on investment, but requires broadcasting programs of a certain set of standards.
In the creative industries, the emphasis is not so much on compensation and most artists are incentivized to work through a utilitarian wealth (Caves, 2003). Artists may choose to live a more bohemian lifestyle by devoting their lives to their work and dismissing the importance of monetary rewards (Eikhof & Haunschild, 2007). In addition, the success of a product is highly uncertain in the creative industries (Caves, 2003). The artists have difficulty predicting if the product will be a success, let alone the managers. Moreover, Caves (2003) discusses a separation of artistic input and humdrum inputs. The humdrum inputs ensure that the artistic work is seen by consumers. But more importantly, the humdrum inputs respond to economic incentives (Caves, 2003).
Thus a combination of high uncertainty in performance and a separation of artistic and humdrum inputs questions whether executive compensation contracts can incentivize managers working in the creative industries. This raises the question: what is the effect of executive compensation on performance in the creative industries?
Performance in the creative industries cannot be measured via standard accounting terms, such as return on investment or return on assets. The government is the sole shareholder and the Public Broadcasting Service does not need to distribute any of its profits to its shareholder. In this research performance in the Public Service Broadcasting is defined as economic and artistic performance. Economic performance is determined by income, whereas artistic performance is defined by awards and market share.
Most of the literature on the creative industries concentrates on consumption and how value is determined (Townley, Beech & McKinlay, 2009). There is also sufficient literature on the cooperation between humdrum and artistic inputs (Caves, 2003, Currid, 2007). In addition, there is literature on management in the creative industries (Bhansing et al., 2012; Peltoniemi, 2015). Executive compensation, which has been researched extensively in the regular business settings (Murphy, 2012), has not yet been studied for the creative industries. Therefore, this paper focuses on whether performance is positively or negatively related to executive compensation in the creative industries, taking into account artistic and economic indicators. The literature on executive compensation is generally inconclusive on whether performance is positively influenced by higher pay, yet in practice pay-for-performance seems to be gaining in popularity (Murphy, 2012).
The research question of this paper is ‘what is the effect of executive compensation on performance in the Public Service Broadcasting?’
An answer to this research question could provide insights for government policies for the creative industries and executive compensation packages. Suppose economic incentives do motivate managers to optimize performance, excessive compensation can appear to be unjustified in comparison to other merit goods such as education and health service. In order to initiate justifying excessive compensation packages in the Public Service Broadcasting, a relation between performance and pay should at least be present.
Additionally, an answer to the research question could provide a useful extension to the existing literature on management theories in the creative industries.
The use of the Dutch and British creative industries could represent a large part of Europe, where it is common to have a mixture of subsidized and non-subsidized broadcasting services.
This quantitative research uses data from the UK and the Netherlands from 2006 until 2013. The British Public Service Broadcasting is the BBC, whereas the Dutch is called the NPO. For each year executive compensation was measured, in addition to subsidies and advertising income, the number of awards won and market share. Performance is divided into economic and artistic performance. Economic performance is defined as income, whereas artistic performance is set up according to the selection system theory, divided in market, peer and expert performance. This quantitative research investigates a relation between performance and executive compensation via a regression analysis.
Following this introduction the literature review provides a discussion of the theoretical background on executive compensation and the creative industries. In addition, the literature review will elaborate on the Public Broadcasting Service and contains the hypotheses and theoretical framework. Subsequently, a chapter on the data and methodology explains in detail the sample and the quantitative methods used. The succeeding chapter presents the results of the quantitative research. A discussion of the results will follow and the thesis closes with a conclusion.
2. Literature review
The executive board consists of a group of managers who run the company on behalf of the shareholders. The manager or the executive is appointed either by the executive board itself or by the supervisory board. In general, a compensation committee is responsible for the executive compensation package. Management literature has extensively studied executive pay. That is, currently there are multiple theories that explain various structures in executive compensation packages, depending on which view is accepted. The Efficient Contracting Theory and the Managerial Power Theory are the most common theories in the research literature on executive compensation. More recent research suggests that these theories do not fully acknowledge tax advantages, legislation and the political climate (Murphy, 2012; Frydman & Jenter, 2010), which created the need for the Perceived Costs theory and the Politics of Pay theory. This research uses the Efficient Contracting Theory as the underlying theory and introduces it to the Public Service Broadcasting, which is a part of the creative industries.
The first section in the literature review describes the existing executive compensation theories. The second section introduces the characteristics of the creative industries. Thirdly, the setting in the Dutch and British Public Service Broadcasting is explained and is linked to general European settings. The literature review continues with the hypotheses, which are categorized according to performance, and separated into economic and artistic performance. Lastly, the conceptual framework ends this chapter to complete the understanding of the topic of this research.
2.1 Executive Compensation
As stated in the preface of this chapter, there are multiple theories on contracting. The first and most common is the Efficient Contracting Theory (Murphy, 2012), also known as the Optimal Contracting Theory (Bebchuk & Fried, 2003). This research uses the Efficient Contracting Theory as underlying theory. Therefore Efficient Contracting Theory is explained extensively, while other theories are explained briefly.
Efficient Contracting Theory explains executive compensation as pay-for-performance with pay-for-performance commonly regarded as maximized profit (Reder, 1947). The Efficient Contracting Theory results from the agency problem.
Agency theory is part of the contractual view identified first by Berle and Means (1932), developed by Coase (1937) and later by Jensen and Meckling (1976). The principal, in this case the shareholders, needs to ensure that the agent, the manager hired to operate in the business, operates in the principal’s best interest. The principal cannot observe the agent’s actions, but can only observe the outcome of the agent’s actions (Grossman & Hart, 1983). The outcome is not only based on the agent’s actions, but involves other factors, which are outside the agent’s control (Grossman & Hart, 1983; Buck, Bruce, Main & Udueni, 2003). Examples of actions of executives that are not in the shareholders’ best interest are entrenchment, empire building or the creation of perks. Adam Smith identified the separation of ownership and control in 1776 as “Being managers rather of other people’s money than of their own, it cannot well be expected, that they should watch over it with the same anxious vigilance with which the partners in a private copartnery frequently watch over their own … Negligence and profusion, therefore, must always prevail, more or less in the management of the affairs of such a company.” (As cited in Murphy, 2012, p. 134).
When there is a conflict between a principal and an agent arising from the separation of ownership and control (Bebchuk and Fried, 2003), a contract can be used in order to align the interest of both parties (Buck et al., 2003). Optimally, a contract is set up to specify what the agent does with the funding provided by the principal and how the profits are divided (Shleifer & Vishny, 1997). In practice, this type of complete contracts are technologically infeasible because the future is unpredictable (Shleifer & Vishny, 1997). Therefore, managerial rewards can be output-contingent so that the interest of the executive is aligned with that of the shareholders, minimizing the agency problem (Shleifer & Vishny, 1977; Grossman & Hart, 1983; Hart & Moore, 1990, Buck et al., 2003). However, when the rewards are more sensitive to output, it puts a higher risk on the executives, who in turn demand to be compensated for additional risk (Conyon & Murphy, 2000).
What outcomes are optimal for the principal depends on the kind of ownership structure. A shareholder in a publicly traded firm benefits from dividend pay. A shareholder such as the government requests a different outcome, most common in the form of the general public’s best interest.
Another theory is the Managerial Power view, which also continues on the agency theory and claims that the executives have the ability to influence the level and the composition of their own compensation packages (Murphy, 2012; Bebchuk & Fried, 2003). Yermack (1997) concluded as one of the first researchers that even though research found a relation between compensation and performance, managers might have foreseen an improvement on performance and requested performance-based compensation. The managers were to sign contracts, which would incentivize them to manage the firm towards a shareholder optimal output, while the managers knew that without interference the output would already be good. Bebchuk and Fried
(2003) attributed to the Managerial Power view, as to claiming that the individuals who set the compensation packages find it easier to go along with the executive than to fight the executive and other friends on the board.
While the first two theories have formed the base for executive compensation theory, these are not consistent with the evidence according to Frydman and Jenter (2010). Murphy (2012) filled the gap with two other theories, namely the Perceived Costs Theory and the Politics of Pay Theory. The Perceived Costs Theory is much less acclaimed by financial economists, due to its suboptimal decision-making. The Perceived Costs Theory states that there is a difference between the economic and perceived costs because of the accounting and tax advantages on stock options used in contracts (Murphy, 2012; Hall & Murphy, 2003). The Politics of Pay (Murphy, 2012) continues on the Efficient Contracting Theory, but explains any remaining factors by counting government intervention often aimed at curbing excessive compensation. This theory explains the boom in stock options in the 1990s and the large decrease in stock options from 2001-2005 (Murphy, 2012).
2.2 Creative Industries
The subject of creative or cultural industries has been introduced as a separate stream of literature. Peltoniemi (2015) has found that the reason for a growing interest in this field is due to multiple aspects. Previous literature has emphasized the importance of selection systems for cultural goods; the inherent uncertainty concerning the success of a product; the definition which entails its low or non-utilitarian value; and the conflicts between the artistically motivated artists and the economically motivated managers (Peltoniemi, 2015). According to Peltoniemi (2015) the definition for creative industries is all organizations that produce
experience goods with considerable creative elements and aim these goods at the consumer market via mass distribution. This section continues with explaining the various aspects of the creative industries. It starts with explaining the definition with respect to value, continues with the selection systems followed by the uncertainty aspect. Furthermore, the conflict between artistic and economic sides of the organization is explained. Lastly, this section ends with a discussion on the use of contracts for executives in creative industries.
The value of a cultural good offers low utilitarian value and a high level of aesthetic or expressive value (Hirsch, 2003). The concept of creative industries as according to Hirsch (1972) consists of three strategies: the deployment of “contact” men to organizational boundaries; overproduction and differential promotion of new items; and the cooptation of mass-media gatekeepers. The concept of “contact” men is related to the existence of individuals responsible for the relations with gatekeepers, people outside the firm who serve as regulators of products to mass media (Hirsch, 1972). Overproduction is common in an industry with low capital investment and high uncertainty on demand (Hirsch, 1972). The costs for pretesting a product are higher than producing more “failures” for each success, which is how overproduction starts (Hirsch, 1972). Lastly, the gatekeepers create a dilemma for the producers since the producers want the gatekeeper to choose their product to be delivered to the mass media, but they also require the gatekeeper to be independent and of high quality (Hirsch, 1972).
The ‘nobody knows’ factor embodies the great uncertainty a producer of a creative good faces about the audience’s perception of the quality (Caves, 2003), which is a problem because of the large sunk cost. An artist must be able to recover from creative goods that resulted in lesser incomes, by capturing enough income from
goods that are beneficial. The value of the creative good also consists of two aspects: the cultural value and the economic value of the product (Caves, 2003). The creative good is taste-driven and depends first on subjective evaluation by gatekeepers and then on consumer’s perspective (Currid, 2007). Determining the cultural (nonnumerical) value of a creative product happens via a social process of legitimization and valorization, and depends on concepts such as the Veblen effect, cultural commodification and the origin of the product (Currid, 2007). When the value of a product is substantially higher or increasing due to exclusiveness, one speaks of the Veblen effect. When artists produce to conform to demands of the critics, they adopt cultural commodification (Currid, 2007). The value of a product can also be determined by the place of origin, for example Hollywood Films and French art (Currid, 2007). The economic value of a creative good, the price, is related to the cultural value and is determined by intangible social value by certain people, can be peers or experts (Currid, 2007).
Art exists when the recipient conceives it as art via individual reactions or aesthetic value (Eikhof & Haunschild, 2007). The recipient can exist in three forms. Selection system theory describes these three forms as the market, peers and experts in the field. In the definition of market as recipient there is a distinction between producers and consumers. Producers are the selected and consumers are the selectors in the selection system theory (Wijnberg & Gemser, 2000). For peer recipients, the selectors and the selected are at a similar position within the industry, for example they are both producers (and possibly both are also consumers) (Wijnberg & Gemser, 2000). In an expert selection system the selectors are external; they are neither consumers nor producers, and have specialized knowledge on the output of the selected (Wijnberg & Gemser, 2000).
In the creative industries there is a constant tension between artistic and economic performance (Peltoniemi, 2015). Most literature on creative industries indicates a clear distinction between an economic and an artistic side within an organization.
First, the artistic side of performance is explained by the production of art for art’s sake (Eikhof & Haunschild, 2007). Art for art’s sake involves the attitudes of artists towards their work, incorporating the utility derived from the production process and the unique style or technique used in the production process (Caves, 2003). Eikhof and Haunschild (2007) discuss the bohemian lifestyle that accompanies the artistic attitude of artists. The bohemian lifestyle defines the devotion of individuals to their work and dismisses the importance of monetary rewards that come with it (Eikhof and Haunschild, 2007). This defines the labor factor in the creative industries, indicating that an economic incentive is less important than aesthetic incentive. Caves (2003) found that artist may accept wages for creative work that fall short of their opportunity cost in humdrum employment. Moreover, the creative industries has a persistent oversupply in labor (Peltoniemi, 2015), which combined with the fact that people in the creative industries would rather work for the utilitarian incentive than for the economic incentive, results in relatively lower incomes.
Second, the economic side of performance is explained by a market value orientation (Eikhof & Haunschild, 2007). In this setting the principal legitimization for producing specific output is the exchange on a market, as opposed to producing for the greater good (Eikhof & Haunschild, 2007). When a product is created purely for an exchange on the market, the creator needs to consider cost efficiencies in order to gain in profits. Most companies in the creative industries rely on government
funding, in the form of subsidies and are not necessarily focused on making a profitable product. On the other hand, public budget cuts have led to more focus on economic efficiency. Managers have not only a focused on the cultural goals, but also achieving them in a cost-efficient way (Eikhof & Haunschild, 2007). Caves (2003) denoted the economic side of a company as humdrum inputs. Humdrum inputs arrange for the artist that consumers get to know the products. However, humdrum agents respond to economic incentives (Caves, 2003).
The creative industries are known for the use of contract in order to address complex incentive problems, such as contracts with star actors (Caves, 2003). Where the creative industries would not immediately come to mind when thinking of sophisticated commercial contracts, the use of revenue-sharing joint ventures with up-front payments and real option contracts are common (Caves, 2003). In some cases the conflict between artistic and economic perspectives within an organization emerge within its top management. In order to establish an optimal contract for the executives, a compensation committee considers what incentivizes the executives. How the contract is constructed depends on what the shareholders value most, but also depends on what the executives value most. The contract considers a tradeoff, i.e. whether the executives should focus more on quality or on financials. The focus can originate in the background of the executive, who has a business or artistic background, and is hired for a specific business or artistic strategy. The firm can consist of dual leadership, with multiple foci due to heterogeneity in the board (Voss, Cable & Voss, 2006; Bhansing, Leenders & Wijnberg, 2012), e.g. the CEO has an artistic perspective, which is complemented by the CFO, who has an economic perspective. According to Peltoniemi (2015) an artist with a pure artistic perspective
teamed up with a professional with a pure profit motive is hardly likely to perform well on either dimension.
2.4 Public Service Broadcasting
Within Europe there are various national policies for the Public Service Broadcasting. Hallin and Macini (2004) categorize the various schemes used in European countries as the ‘liberal model’, the ‘polarized pluralist model’ and the ‘democratic corporatist model’ (Bardoel & d’Haenens, 2008). The liberal model, also called the North Atlantic model, is typical for the United Kingdom and its former colonies. The Public Service Broadcasting in these countries remains without interference from the government and relies on self-regulation and internal control (Bardoel & d’Haenens, 2008). The polarized pluralist model is found in the Mediterranean countries. In this model, there are considerable levels of government intervention and politicization (Bardoel & d’Haenens, 2008). Lastly, the democratic corporatist model, also called the Northern European model, is present in the Netherlands, Austria, Switzerland, Belgium, Germany and the Scandinavian countries. In these countries there is also government intervention, but with the protection for press freedom (Hallin & Macini, 2004). This research tries to focus on the European Union. However, due to the availability of data, only the UK and the Netherlands are researched. Because of similarities in the systems, the results can be used for most European countries, except for the Mediterranean countries.
The Public Service Broadcasting in the UK uses license fees as income (Bardoel & d’Haenens, 2008). These license fees are gathered by the Public Service Broadcasting from the public, and are obligatory for every household with a television broadcasting live to the public. Since the BBC gathers directly from the public, there
is no interference from the government. Additionally, the BBC is not allowed to sell advertising time or receive sponsorships in order to run purely for the general public interest (BBC.co.uk, 2015). The BBC also operates outside of the UK, and is allowed to sell advertising time for broadcasting in the rest of the world.
The Public Service Broadcasting purposes in the UK are as follows: the Public Service Broadcasting needs to inform the public and increase its understanding of the world, to stimulate knowledge and learning, to reflect the UK cultural identity, to represent diversity and alternative viewpoints (Ofcom, 2013).
In the Netherlands, the government does not allow profit maximization for the Public Service Broadcasting and prohibits commercial breaks during programs; the Public Service Broadcasting is subsidized in order to be able to breakeven. The commercial breaks in between programs are managed via a separate company, the Ster3, in order to ensure the NPO operates in the general public’s interest instead of the advertisers’ interest. The subsidies the NPO receives include the advertising revenues assembled by Ster. The Public Service Broadcasting is allowed to have some surplus on its accounts, but will not distribute any profits to its shareholder, the government. This surplus is therefore not a goal of the Public Service Broadcasting. Three topics are the main focus of its shareholder, the Dutch government. These are information, culture and education (Commissariaat voor de Media, 2013).
In the Netherlands the NPO consists of multiple smaller broadcasting associations. These associations are supposed to have members and based on the number of members the government decides if the association should continue to exist. The broadcasting associations thus needs viewers, who can be persuaded to become members.
3
Meijer (2005) defines viewers in three categories, namely the ‘citizen’, ‘consumer’ and the ‘enjoyer’. First, the citizen category is defined as socially aware and helps to benefit the society as a whole (Hastings, 2004). A program type matching the citizen would be a program that encourages viewers to be democratically involved, creates a community feeling, educates and informs (Meijer, 2005). Second, the consumer category is established as target groups for the advertisers. By addressing, creating and delivering markets the program is much more attractive for advertisers (Meijer, 2005). Lastly, the enjoyer wants to be entertained. A program type matching the enjoyer would be a program that gives visual pleasure, charms the viewer and creates utopian escapism (Meijer, 2005).
Meijer (2005) explains how the Public Service Broadcasting should focus on consumer impact, in order to better be able to achieve the three main objectives set by the Dutch government. The three main objectives given to the Public Service Broadcasting by the government are to provide television programs that are informative, cultural and educational, but aside from that the Public Service Broadcasting should broadcast programs that focus on youth, politics and sports (Rijksoverheid, 2015b). Meijer (2005) argues that the Public Service Broadcasting must also focus on the consumer as an ‘enjoyer’ and broadcast programs that specifically entertain the consumer. This focus would lead to a larger audience reach and market share. However, more recently State Secretary for Education, Culture and Science, Sander Dekker has claimed that the Public Service Broadcasting needs to concentrate on diversification and less on amusement (Rijksoverheid, 2015b). The focus of the government can change with each policymaker. The Public Service Broadcasting should retain its number one market position, which it still holds in most European countries, and so forth justify their existence as public service institution.
The policy of the government concentrates most on the citizen, whereas the advertisers focus on the consumer and the audience relates most to the enjoyer. The Public Service Broadcasting should concentrate on the three types of viewers since subsidies also depend on audience, because the broadcaster needs viewers, and next to subsidies, advertising is also contributes to income.
2.5 Hypotheses
In order to operationalize performance this research makes a distinction between economic and artistic performance. Economic performance concentrates on financials and comprises all elements contributing to additional income. Artistic performance focuses on the artistic quality and comprises of aspects that contribute to artistic excellence. This section starts by explaining how economic performance is defined and continues with the first hypotheses. Subsequently, artistic performance is construed, followed by one more hypothesis.
2.5.1 Economic Performance
Profit maximization is not allowed to be a goal in the Public Service Broadcasting, a goal of the managers is thus to seek continuity; being able to sustain receiving subsidies, sponsorships and income from commercials to make the broadcasting service viable and to be able to deliver on their strategic goals.
In the previous literature on efficient contracting, compensation is set to incentivize the executives to create a highest possible return on investment for the shareholders (Murphy, 2012; Bebchuk and Fried, 2009; Shleifer and Vishny, 1997). Previous research has found positive pay-performance sensitivities, such as Murphy (1985), Coughlan and Schmidt (1985) and Jensen and Murphy (1990).
Murphy (1985) found a strong relationship between pay and performance in publicly traded Fortune 500 companies. Coughlan and Schmidt (1985) found a positive pay-performance relation in companies listed in Forbes magazine. Jensen and Murphy (1990) found a small positive relationship, but also report on a declining relationship over the years. These researches are to some extent outdated and mostly based on US data, more recent research reports on different results. Conyon and Murphy (2000) and Buck et al. (2003) performed studies in the UK. They concluded that there is a pay-performance sensitivity, though relatively small. Haid and Yurtoglu (2006) reported low pay-performance sensitivities in Germany. Duffhues and Kabir (2008) researched the pay-performance sensitivity in the Netherlands and concluded to find no evidence for a pay-performance relation. There is no consensus on whether efficient contracting actually solves the agency problem. Furthermore, it seems the use of this type of contracting is currently rather small in the UK and the Netherlands. However, the focus of the government, the shareholder of the Public Service Broadcasting, is not return on investment. The government wants to ensure quality television for the general public. The government provides subsidies, in the form of direct subsidizing or a license fee. A second possible source of income for the Public Service Broadcasting is advertising revenues. As explained in previous sections, the NPO receives advertising income via an external organization, whereas the BBC does not broadcast commercials and receives no income via advertising. Lastly, another source of income is sponsoring. The Public Service Broadcasting can have sponsors for cultural programs or for sport related programs. Sponsorship is considerably low and is not considered as a significant contributor to income or performance in the Public Service Broadcasting, and is not in the scope of this research.
Income for the Public Service Broadcasting in the UK is generated via a license fee. Every household needs to pay the fee in order to receive the BBC channels on their televisions. The license fee is a stable and predictable source of income for the Public Service Broadcaster; there is no interference from the government and there is a direct link between the public and the broadcaster (European Broadcasting Union, 2015). A decrease in quality or popularity of the broadcaster would mean the public might choose for commercial broadcaster and unsubscribe to the public service broadcaster. On the other hand, if performance of the Public Service Broadcaster goes up, audience increases, resulting in an increase in income.
Income for the Public Service Broadcasting in the Netherlands is generated via the direct funding from the state budget. This type of income has the disadvantage of a high reliance on political interference and pressure, which can lead to budget cuts (European Broadcasting Union, 2015). After 2000, when the system changed from license fee to direct funding, the government immediately posed large budget cuts for the Public Service Broadcasting (European Broadcasting Union, 2015). The executives of the broadcaster may try to persuade the government not to cut budgets, but will only successfully convince them if there is proof that the broadcaster is providing quality television to a significantly large part of the population. Therefore an increase in performance will either lead the government to decide to not cut or increase the subsidies. Moreover, with a larger reach more income from advertising will be generated. If the executive is able to improve performance so that the audience reach increases, advertising income will also increase.
H1a: A higher executive compensation results in an increase in income of the organization.
H1b: A higher executive compensation results in no decrease in income for the organization.
2.5.2 Artistic performance
Artistic performance of the Public Service Broadcasting can be measured by various aspects. This research operationalizes artistic performance by taking into account selection system theory, where the selectors determine value and outcome of competitive processes (Bhansing et al., 2012). Selection system theory identifies three types of selectors, market selectors, peer selectors and expert selectors (Bhansing et al., 2012). In a market selection system there is a distinction between producers and consumers. Producers are the selected and consumers are the selectors (Wijnberg & Gemser, 2000). Artistic performance as measured by market selectors is in this research measured via market share. In a peer selection system the selectors and the selected are at a similar position within the industry, this means they are both producers (and possibly also consumers) (Wijnberg & Gemser, 2000). In an expert selection system the selectors are external; they are neither consumers nor producers, and have specialized knowledge on the output of the selected (Wijnberg & Gemser, 2000). In this research peer selection is measured in peer-jury awards, whereas expert selection is measured in expert-jury awards.
If the broadcaster performs better, one of the three selection systems must increase. When market performance increases, the broadcaster will capture a larger part of the population in their audience reach. An increase in audience reach should be better for incomes, advertising or subsidies. Moreover, increase in peer or expert
Higher executive compensation leads
to overall better performance for firms in the creative
industries Economic Performance H1: higher executive compensation leads to an equal or higher income economic performance Compensation Artistic Performance H2: higher executive compensation leads to higher artistic performance Compensation
performance is better for the reputation of the executive. Especially in the artistic environments, peer opinions are highly valued (Caves, 2003; Bhansing, Leenders and Wijnberg, 2014).
This suggests the following hypothesis:
H2: A higher executive compensation will associate with higher artistic performance.
2.6 Conceptual Framework
Management in the creative industries can be better understood when the motivations of managers are discovered. Therefore, testing the hypotheses will lead to an indication of whether monetary incentives work on a management level in the creative industries. Currently, there is a strong belief for pay-for-performance in the general business world, through the efficient contracting perspective, which could continue on into the creative industries. Figure 1 shows the conceptual framework for this research.
3. Data and Methodology
This chapter describes the sample, data collection and the methodology. It provides a clear description of the variables and discusses the descriptive statistics that give an initial look upon the data set. The data and methodology chapter ends with correlation matrices for both the Dutch and British variables.
3.1 Sample
The sample is selected based on availability. In the Netherlands, executive compensation is available from 2006-2013 due to the law WOPT, which initiated in 2006. For the UK the same sample period is used in order to be able to compare the outcomes. Table 1 summarizes all the variables and shows their sources.
Compensation of each executive, in other words, each board member, is taken into account as the (only) independent variable. The primary source of executive compensation data is the annual report of the Public Service Broadcasting of the Netherlands and the UK. Both the NPO and the BBC are required to report the compensation of the executive board members, in the Netherlands under the law WOPT1, in the UK under the Companies Act 20064.
Performance is the dependent variable and is measured from two different viewpoints, in economic and artistic performance. Economic performance is measured from the level of income, through subsidies. These values are also disclosed in the annual report. In addition to subsidies, economic performance for the NPO is also measured through advertising revenues. The BBC is prohibited from advertising; therefore advertising revenue is only used for the Dutch Public Service Broadcasting. The subsidies also include the advertising revenues, since the Dutch government
collects advertising revenues from a separate firm, the Ster3. The Ster is responsible for the sale of commercial time for the Dutch Public Service Broadcasting.
Secondly, artistic performance is measured based on the selection system theory. Selection system theory indicates that specific selectors, expert, peer or market selectors, determine the value of the output in creative industries (Bhansing et al., 2012).
In a peer selection system the selectors and the selected are at a similar position within the industry, hence they are both producers (and possibly also consumers) (Wijnberg & Gemser, 2000). Peer selection is measured in peer-jury awards, in this case the ‘Gouden Kalf’ in the Netherlands and the ‘BAFTA’ awards in the UK.
In addition, expert reviews characterize artistic performance, for example when an expert appraises the output of the Public Service Broadcasting as quality material, one can assume good artistic performance. In an expert selection system the selectors are external; they are neither consumers nor producers, and have specialized knowledge on the output of the selected (Wijnberg & Gemser, 2000). Expert selection is measured in expert-jury awards, in this case the ‘Zilveren Nipkow’ in the Netherlands and the ‘National Television Awards’ in the UK.
In a market selection system there is a distinction between producers and consumers. Producers are the selected and consumers are the selectors (Wijnberg & Gemser, 2000). Artistic performance as measured by market selectors is measured via market share. The SKO5, the television audience measurement service of the Netherlands publishes the official television audience figures of the Dutch population (SKO, 2015). Ofcom, the telecommunications regulator in the UK publishes the
Table 1. List of variables
This table displays information on each variable, such as a description and the main source for both the UK and Dutch data. References and links for the sources can be found in chapter 7. Reference list.
Variable Description
Independent Variable
Compensation Executive compensation as measured in local currency
Source: annual reports of 2006-2013 of the NPO (NL) annual reports of 2006-2013 of the BBC (UK)
Dependent Variables
Peer Performance measured via the peer selection system, i.e. peer-jury awards. For the Netherlands: Gouden Kalf. For the UK: BAFTA. See Exhibit 1 in the appendix
Source: Winners of 2006-2013 of the Gouden Kalf (NL)
Winners of 2006-2013 of BAFTA (UK)
Expert Performance measured via the expert selection system, i.e. expert-jury awards. For the Netherlands: Zilveren Nipkow. For the UK: Royal Television Society Programme Awards.
See Exhibit 1 in the appendix
Source: Winners of 2006-2013 of Zilveren Nipkow (NL)
Winners of 2006-2013 of RTS (UK)
Market Performance measured via the market selection system, i.e. market share. For the Netherlands: the Stichting Kijkonderzoek. For the UK: Ofcom.
Source: annual reports of 2006-2013 of the SKO (NL)
annual reports of 2006-2013 of Ofcom (UK)
Subsidies Performance measured via subsidies in local currency
Source: annual reports of 2006-2013 of the NPO (NL)
annual reports of 2006-2013 of the BBC (UK)
Advertising (NL only) Performance measured via advertising in local currency
Source: annual reports of 2006-2013 of the NPO (NL) Control Variables
Economic Growth GDP growth in the country
Source: Worldbank
Public Policy Governmental policy changes towards the Public Service Broadcasting
Source: Mediamonitor (NL)Ofcom (UK) Large Events Number of large (sport) events in that year
Source: Mediamonitor (NL and UK) Internet Usage Number of internet users
official television audience figures of the British population (Ofcom, 2015).
Table 1 also discloses the control variables used in the regression. These control variables are added to the research because they affect the dependent variable without influence of the independent variable. First, GDP growth is added. GDP growth is expected to affect performance, because this might increase subsidies. Second, public policy is also added, a change in public policy on Public Service Broadcasting will certainly affect performance without any interference of the executives. Third, large (sports) events is added as a control variable, because not only do these events attract viewers, but usually additional funding is received from the government to finance the broadcasting of these particular events. The large (sports) events considered in this research are the Olympics for both the UK and the Netherlands and the Royal wedding in the UK. Lastly, the number of internet users is added as a control variable. Both the BBC and the NPO have the option of online watching, with more people having access to the internet, a larger possible audience can be captured.
The next section shows how these variables are researched in a quantitative manner, via a regression analysis.
3.2 Regression Analysis
From the annual reports a panel data set is gathered manually and a regression analysis is performed as used in Murphy (1985). The research question investigates the effect of executive compensation on performance. A linear regression is suited, with a dependent variable and one independent variable. The general form of a linear regression is:
Concerning the research question, the Y variable, the dependent variable is performance. The independent variable X is in this research executive compensation. Since the linear regression defines a line in a graph, β0 is defined as the intercept. β0 is
the value of Y where X is 0. The error term e, also known as residual or remainder, contains all other factors influencing the dependent variable. β1 measures the effect of
X on Y, in this research this provides the answer to the research question and is the effect of compensation on performance. Equation (1) has no control variables included. Control variables are added to the regression in order to extract the effect of these factors, which also influence the dependent variable. Furthermore, the sample data is gathered over a period of time, from 2006 until 2013. The data is said to be time series data and requires a time series analysis. In the regression ‘t’ defines the year. The appropriate linear regression for this research is:
Performancet= β0+ β1 ln (Compensation)t+ β2 GDP growtht+ β3 Public Policyt
+ β4 Large Eventst+ β5 Internet Userst + et (2)
Compensation entails the total compensation received per executive per year. Performance is measured in economic and artistic performance. Economic performance as measured in the amount of subsidies received per year and the amount of advertising income generated per year. Artistic performance is measured by market, peer and expert performance. Market performance is defined by the market share of the Public Broadcasting Service of year t. Peer and expert performance is defined by the number of awards won in year t. The control variables are the same for each type of performance. The following equations are the appropriate linear regressions for the different types of performance:
Ln (Subsidies)t = β0 + β1 ln (Compensation) t + β2 GDP growtht + β3 Public
Policyt + β4 Large Eventst+ β5 Internet Userst + et (3.1)
Ln (Advertising)t = β0+ β1 ln (Compensation) t + β2 GDP growtht + β3 Public
Policyt+ β4 Large Eventst+ β5 Internet Userst + et (3.2)
Market Sharet = β0 + β1 ln (Compensation) t + β2 GDP growtht + β3 Public
Policyt + β4 Large Eventst+ β5 Internet Userst + et (3.3)
Peer Performancet= β0 + β1 ln (Compensation) t+ β2 GDP growtht+ β3 Public
Policyt+ β4 Large Eventst+ β5 Internet Userst + et (3.4)
Expert Performancet = β0 + β1 ln (Compensation) t + β2 GDP growtht + β3
Public Policyt+ β4 Large Eventst+ β5 Internet Userst + et (3.5)
The analysis delivers on an estimated β1, which is the estimated effect of
compensation on performance. Furthermore, the research looks at the change in executive compensation and performance, thus if the government were to grant higher packages, would this increase performance. Logarithmic data also reduces skewness of the size distribution of the sample executives and yields easily interpretable regression coefficient (Murphy, 1985).
Control variables are added to the analysis to control for other factors influencing performance. Factors influencing the performance of the firm are for instance GDP growth, public policy, sport events and the increase in internet usage
for television watching. Public policy changes and sport events are used as dummy variables: the value will be 0 if there is no change in policy or large sport event, 1 if there is. β2,β3,β4,and β5 are estimators of the effects of the control variables.
This research uses the statistical software package STATA. STATA computed the descriptive statistics and the correlations matrices for the next sections and regression results of the succeeding chapter.
3.3 Descriptive Statistics
In this section descriptive statistics are displayed in order to present a first view upon the data. These statistics describe the data set. Per Public Service Broadcasting the variables are presented with the number of observations. Next to that the mean, which is the average, and the standard deviation, which is the spread, are given for each variable. Furthermore, the lowest and highest values for each variable are shown as minimum and maximum.
Table 2 presents the descriptive statistics for the Dutch Public Service Broadcasting, the NPO. First, average compensation lies at € 232,160 per executive. This amount is above the limit of the law (WNT), which is set the salary of a minister at € 228.599 (Rijksoverheid, 2015a). The Dutch compensation law (WNT) was enforced in 2013, indicating that on average future executive compensation needs to be lowered.
Performance is separated into artistic and economic performance. Artistic performance is measured via selection system theory as explained in the previous section. Market selection, measured via the market share, has an average of 35.66%.
Table 2. Descriptive Statistics of the NPO
This table displays summary statistics for every variable of the Dutch Public Service Broadcasting the NPO. Compensation is displayed in thousands. Subsidies and Advertising are displayed in
millions.
Variable Obs Mean Std. Dev. Min Max
Compensation 31 232,160 41,740 130,600 304,900 Ln(compensation) 12.324 0.210 11.780 12.628 Performance Market 35.658 1.663 33.200 37.700 Peer 1.00 - 1.00 1.00 Expert 2.00 - 2.00 2.00 Subsidies (in 000s) 719,010 50,570 600,400 755,500 Ln(subsidies) 20.437 0.090 20.217 20.499 Advertising (in 000s) 200,750 12,190 188,000 220,000 Ln(advertising) 19.115 0.058 19.052 19.209 Control Variables GDP Growth 1.032 2.401 -3.300 4.200 Policy change 0.226 0.425 0.00 1.00 Large Events 0.484 0.508 0.00 1.00 Internet Users 89.445 3.571 83.700 94.000
Source: author’s own calculations
This means that on average over the years 2006-2013, the NPO had a market share of 35.66% on the Dutch audience market. The remaining share is captured by commercial broadcasting stations. Peer selection has a mean of 1, stating that the average number of awards won is 1. However, looking at the minimum and maximum, the number of awards won according to the peer selection does not differ from 1. Likewise, the other type of awards, the expert-jury awards displays no variation; the mean, maximum and minimum are 2. This means that every year, from 2006 until 2013, the NPO has won one Gouden Kalf award and two Zilveren Nipkow awards. Because the lack of variation in peer and expert performance causes problems for the regression analysis in the next chapter, a further investigation to other types of
performance measurement possibilities was performed. However, there were no other peer or expert awards for the full sample period available6.
Economic performance is measured in subsidies and advertising. Subsidies are on average € 719,010,000. Advertising has a mean of € 200,750,000. All three variables measured in currency are also displayed as converted to natural logarithms. For the regression, in the next chapter, only the logarithmic values are used as explained in the previous section. Additionally, the descriptive statistics of the control variables are added. The average GDP growth in the Netherlands for the period 2006-2013 is 1.03%. The variables Policy Change and Large Events are dummy variables, with a binominal value of 0 and 1. The value of the variable is 0 when there is no policy change or large event, respectively, and 1 if there is. The average number of internet users per 100 people is 89 persons.
Furthermore, in order to analyze compensation more thoroughly, a table and graph are presented. Table 3 shows the amount of compensation per board member. The table shows that average compensation increased until 2010, after which it declined in 2013, due to the lower compensation of newer board members. In contrast, the compensation of the CEO actually increased in 2013.
Graph 1 shows the total amount of compensation for the entire board per year for the Dutch Public Service Broadcasting. It shows in 2008 compensation was much higher due to the replacement of the CEO. After 2008, total compensation was lowered, however, the board size also decreased by one executive, the manager of the NOS was no longer part of the board. In 2013 the size of the board doubled, which led to a higher total compensation.
6
Based on the hypotheses in chapter 2, performance should be increasing until 2010 and then start to decrease. In order to give an initial visual understanding of the relation between compensation and performance, the performance data is also graphically displayed. The graphs can be found in appendix 2. Based on these graphs, it could be interpreted with prudence that the decrease in (average) compensation after 2010 also relates to the decrease in market performance. Economic performance, consisting of subsidies and advertising, appears to contain little variation and needs to be further evaluated through the regression analysis.
In the next chapter, the results of the regression analysis show how these variations in compensation relate to fluctuations in the performance of the Public Service Broadcasting. Based on the descriptive statistics equations 3.4 and 3.5 will not be regressed for the Dutch Public Service Broadcasting. Without variation a regression analysis cannot be performed. The results of regressions 3.1, 3.2 and 3.3 are shown in the subsequent chapter.
Table 3. Compensation per Board Member of the NPO
This table displays the compensation of each board member per year of the Public Service Broadcasting of the Netherlands, the NPO. The amounts are in Euros.
Year CEO Executive Executive Manager Video Manager Radio Manager D&U Manager Finance Manager
NOS Total Average 2006 271,300 228,500 247,000 181,700 928,500 232,125 2007 248,900 304,900 237,400 186,800 978,000 244,500 2008 159,100 / 146,900 284,900 247,100 193,400 725,400 241,800 2009 224,600 296,400 249,800 768,200 256,067 2010 230,200 293,800 253,500 777,500 259,167 2011 232,900 230,200 296,400 759,500 253,167 2012 238,900 222,500 232,900 694,300 231,433 2013 239,200 229,600 213,400 168,700 178,300 130,600 1,159,800 193,300 Source: NPO Annual Reports
Graph 1. Compensation of the NPO
This graph displays the compensation of each board member stacked per year for the board of the Public Service Broadcasting of the Netherlands, the NPO. The amounts are in Euros.
Table 4 presents the descriptive statistics for the Public Service Broadcasting of the UK, the BBC. Executive compensation at the BBC averages at £ 405,310. The level of executive compensation is much higher in the UK than in the Netherlands, after converting British Pounds to Euros. Artistic performance is separated into peer, expert and market selection. Awards won according the peer selection, which are the BAFTA awards, averages at 12 awards won per year. Performance according to expert selection, which are the Royal Television Society awards averages at 14 awards won per year. Performance measured via market selection, which is measured in market share, averages at 57.68%. Economic performance measured in subsidies
0 200,000 400,000 600,000 800,000 1,000,000 1,200,000 1,400,000 2013 2012 2011 2010 2009 2008 2007 2006
CEO Executive Manager Video (2013 only)
Manager Radio (2013 only) Manager D&U (2013 only Manager Finance (2013 only)
Table 4. Descriptive Statistics of the BBC
This table displays summary statistics for every variable of the Public Service Broadcasting of the UK, the BBC. Compensation and Subsidies are displayed in thousands.
Variable Obs Mean Std. Dev. Min Max
Compensation 78 405,310 174,970 10,000 860,000 Ln(compensation) 12.759 0.716 9.210 13.665 Performance Peer 11,63 2,67 8,00 15,00 Expert 13,50 3,82 9,00 18,00 Market 57,68 5,64 51,10 6,80 Subsidies (in 000s) 3,464,250 133,550 3,242,000 3,656,000 Ln(subsidies) 21.963 0.035 21.899 22.019 Control Variables GDP Growth 0.732 2.288 -4.3 3 Policy change 0.512 0.503 0 1 Large Events 0.641 0.482 0 1 Internet Users 81.344 6.540 68.8 89.8
Source: author’s own calculations
averages at £ 3,464,250,000. Both variables in currency, which are executive compensation and subsidies, are converted to logarithmic values for use in the regression analysis.
Next to the descriptive statistics of the dependent and independent variables, the descriptive statistics of the control variables are added as well. GDP growth in the UK for the period 2006-2013 was on average 0.73%. Policy change and Large events are dummy variables. The variable ‘Large Events’ does not only represent sport events for the UK. The BBC broadcasted the Royal wedding of Prince William andCatherine Middleton in 2011. For this year, the dummy variable was also set at 1. The average number of internet users per 100 people was 81 in the UK.
Furthermore, to understand compensation of the BBC more thoroughly, a table and graph are presented. Graph 2 shows the total amount of compensation for the entire board per year for the British Public Service Broadcasting. Table 5
amount of board compensation seems to fluctuate, with large decreases in 2011 and 2013. The annual reports state a new executive pay policy from 2011 until 2015, which lowers the number of board members and freezes bonuses. However, in 2012 there were two board members receiving large compensation due to loss of office. Not only was there more compensation granted, there were also new board members added, which even further increased the total amount of compensation.
Moreover, the graph shows stacked compensation of the executives per year. For the BBC only the CEO and CFO are specified by title, since the other executive positions varied with title and function over the years. The compensation of the CEO and the CFO show a steady decrease after 2009. From 2011, this is due to the policy change and the combined compensation drops below £ 1,000,000. Table 5 shows the policy change in 2011 very clearly. The number of board members is reduced, as well as the average compensation.
The hypotheses in chapter 2 predict that performance should decrease as of 2011, because of the (negative) change in pay policy. A visualization of the performance of the BBC can be found in appendix 2. The graphs show that market performance is indeed declining, which could predict a relation between compensation and market performance. The other two artistic performance measures and the remaining economic performance measure do not seem to vary in accordance with compensation. However, these interpretations are very limited since none of the control variables are taken into account yet. The regression analysis provides the ability to give more complete interpretations.
The results of the regression analysis in the next chapter show whether the variation in compensation relates to the variation in performance. The regressions performed for the British Public Service Broadcasting are defined as equation 3.1,
3.3, 3.4 and 3.5. Only the regression of equation 3.2 cannot be performed, since advertising is prohibited on the BBC.
Graph 2. Compensation of the BBC
This graph displays the compensation of each board member stacked per year for the board of the Public Service Broadcasting of the UK, the BBC. The amounts are in Pounds.
1,000,000 2,000,000 3,000,000 4,000,000 5,000,000 6,000,000 2013 2012 2011 2010 2009 2008 2007 2006
CEO CFO Executive Executive Executive Executive Executive Executive Executive Executive Executive
Table 5. Compensation per Board Member of the BBC
This table displays the compensation of each board member per year of the Public Service Broadcasting of the UK, the BBC. The amounts are in Pounds.
Year CEO CFO Executive Executive Executive Executive Executive Executive Executive Executive Executive Total Average
2006 788,000 386,000 361,000 329,000 433,000 437,000 25,000 406,000 359,000 268,000 3,792,000 379,200 2007 816,000 440,000 440,000 486,000 419,000 536,000 513,000 413,000 466,000 431,000 4,960,000 496,000 2008 834,000 429,000 413,000 480,000 248,000 515,000 485,000 403,000 124,000 274,000 396,000 4,601,000 418,273 2009 838,000 434,000 419,000 193,000 348,000 328,000 345,000 517,000 488,000 452,000 407,000 4,769,000 433,545 2010 779,000 403,000 385,000 367,000 254,000 440,000 447,000 414,000 436,000 26,000 422,000 4,373,000 397,545 2011 622,000 337,000 328,000 309,000 354,000 349,000 261,000 2,560,000 365,714 2012 503,000 366,000 860,000 181,000 455,000 408,000 78,000 802,000 172,000 10,000 300,000 4,135,000 375,909 2013 467,000 304,000 317,000 378,000 414,000 327,000 217,000 2,424,000 346,286
3.4 Correlation Matrices
In this section correlation matrices are displayed in order to see any first relations between the variables. Per Public Service Broadcasting the variables are displayed with the number of observations. Next to that, the mean, which is the average, and the standard deviation, which is the spread, are given. Lastly, the correlation between each of the variables is displayed. Strong correlations are printed in bold.
Table 6 displays the correlation matrix for the Dutch Public Service Broadcasting, the NPO. Peer and expert performance are not included in the correlation matrix. As seen in the descriptive statistics, there was no variation in the number of awards won by the NPO, therefore they are displayed separately below table 6. Table 7 displays the correlation matrix for the British Public Service Broadcasting, the BBC.
Table 6. Correlation Matrix NPO
This table displays all variables with their means, standard deviation and correlation. Bold values display strong correlations.
Mean SD 1 2 3 4 5 6 7 8 1 Ln(compensation) 12.32 0.21 1.00 2 Market Performance 35.66 1.66 -0.03 1.00 3 Ln(subsidies) 20.44 0.09 -0.05 0.40 1.00 4 Ln(advertising) 19.12 0.06 -0.12 0.83 0.12 1.00 5 GDP growth 1.03 2.40 0.05 -0.57 -0.60 -0.35 1.00 6 Policy Changes 0.23 0.43 0.25 -0.19 0.11 -0.41 0.42 1.00 7 Large Events 0.48 0.51 -0.01 0.53 -0.45 0.57 0.24 -0.06 1.00 8 Internet Users 89.45 3.57 -0.15 0.36 0.80 0.01 -0.75 -0.24 -0.35 1.00 Peer Performance 1 0 Expert Performance 2 0
Table 7. Correlation Matrix BBC
This table displays all variables with their means, standard deviation and correlation. Bold values display strong correlations.
Mean SD 1 2 3 4 5 6 7 8 9 1 Ln(compensation) 12.76 0.72 1.00 2 Peer performance 11.36 2.47 -0.21 1.00 3 Expert performance 13.91 3.47 -0.13 -0.03 1.00 4 Market performance 58.02 5.21 0.12 -0.59 0.20 1.00 5 Ln(subsidies) 21.96 0.04 -0.08 0.49 -0.18 -0.94 1.00 6 GDP growth 0.73 2.29 -0.06 0.53 -0.08 0.20 -0.27 1.00 7 Policy changes 0.51 0.50 0.08 -0.39 -0.32 -0.15 -0.03 -0.54 1.00 8 Large Events 0.64 0.48 -0.19 0.38 0.16 -0.03 -0.25 0.35 0.18 1.00 9 Internet Users 81.34 6.54 -0.09 0.45 -0.22 -0.98 0.95 -0.32 0.24 -0.09 1.00
Source: author’s own calculations
Most noticeable in both tables is that correlation between compensation and performance is negative for all, except market performance in the UK (Table 7). This already seems to contradict the hypotheses, since it was expected that when compensation increases, performance increases as well. The correlations between compensation and performance are weak to very weak according to the correlation matrix, all correlations are below 0.25 in absolute terms.
In table 6, the correlation between executive compensation and market performance is -0.03. The correlation between executive compensation and subsidies and advertising is -0.05 and -0.12, respectively. All correlations are negative and weak, which indicates that the rise of one of the independent variables correlates with the small, fall of the dependent variable.
In table 7, the correlation between executive compensation and peer and expert performance is -0.21 and -0.13, respectively. The correlation between executive compensation and subsidies is also negative, namely -0.08. These correlations, and the correlations of table 6 oppose the hypotheses. The correlation
between executive compensation and market performance is the only positive correlation between the dependent and independent variables, with a value of 0.12.
