Market Concentration, Excess Capacity and
Entry-deterring Behavior in the Dutch Cinema Market
Bachelor Thesis
Faculty of Economics and Business
Specialization: Economics
M.C.M. Silalahi
10621652
Supervisor: M.O. Hoyer
29 January 2015
Table of Contents
Introduction………...3
Literature Review……….…………..5
Data and Methodology ………9
Limitations and Results ………...………15
Discussion………...……..19
Conclusion………..…………..22
Introduction
One’s surrounding when watching movies influences the satisfaction one gains from it. Where, when, and how one watches a film plays a role in determining the overall satisfaction from each movie. In cinemas, it is all about the seating. Some people find sitting in the center of the back row heighten the experience because they get the best view in the auditorium, others may have different opinion on the matter. There are also external factors such as fellow audience. For example: a teenage girl may find it really irritating to be seated next to a gravely obese man with blaring breath and mannerless popcorn eating behavior thus loathing the film experience. In parallel to that, a grown man might find himself in an awkward situation and unable to fully concentrate on the movie after realizing that he inadvertently walked in into a ladies night. It is all very subjective but one often finds oneself wishing to be the only person in the cinema.
Naturally, cinema owners would not wish on this. They have a business to run and in most cases it is not necessarily desirable to have a lot of empty seats in the house. Nevertheless, cinemas may decide to hold certain number of excess capacity though not for quite the same reason.
Excess capacity has been, in numerous theoretical literature been considered as a mean of entry deterrence. Most notably by Dixit (1980), which pointed out that dominant firms (incumbent in the market) might exploit their position in the interest of entry-deterrence by setting high capacity, such that it overwhelms the competition and scare out entrants. Having that said, there are also alternative explanations on why a firm might decide to hold certain level of idle/excess capacity. This make entry-deterrence notion rather contestable.
There have not been many studies that come up with a definite empirical solution to distinguish the motivations for excess capacity. One study by Conlin and Kadiyali (2006) tried to indirectly test whether firms invest in entry deterring capacity according to their market concentration and how market presence affects firms’ incentives to deter entry. Although unable to definitely
disregard the alternative explanation, this study managed to show that more concentrated market have greater idle capacity and that firms with larger market share and thus higher market presence possess more idle capacity.
In this thesis, I would like to examine the case of cinema industry in the Netherlands and whether excess capacity is related to entry-deterring behavior by attempting to partially replicate the empirical approach on how market share affects the excess capacity. The main questions this thesis set out to answer is: how concentrated is the cinema industry in the Netherlands and how does it corresponds in relation to idle/excess capacity. In addition to that, this thesis would also like to explore whether the excess capacity in the industry justifiably constitutes as entry deterrence strategy or seems to satisfy other alternative explanations.
The dataset used to examine market concentration and idle capacity in this thesis subject consists of annual information between 2011-2013 about the number of cinema seats in the Netherlands. The data for each cinema consists of the information about number of seats and screen numbers, location and categorized ownership. Calculating cinema capacity is advantageous because of its simplicity and the static nature of the capacity (it serves only its designated geographical market and are not transportable). Later I would focus the geographical capacity on provincial basis, as to draw a more vivid line of separation between markets in the Netherlands.
The rest of this thesis is organized as follows: the second section of this thesis will try to examine the existing literature on the subject of entry deterring capacity and market concentration. The third section will be the description of the cinema data of in the Netherlands that are successfully acquired for the purpose of the thesis and descriptive statistics resulting from the observation. The fourth section presents the result of the specification attempted to demonstrate the relationship between market concentrations and idle capacity, as well as limitations. The fifth section will discuss alternative explanations for the result. The fifth and last section concludes the analysis on the subject.
Literature Review
Existing theoretical about excess capacity has painted a strong idea about how holding capacity that surpasses its production requirements for dominant firms is an attempt to set up entry barriers. This notion is stressed by Spence (1977), which stated that in an industry, entry is deterred when incumbent firms have sufficient capacity that it makes entrants unprofitable.
Inherently, most of the published paper including Spence (1977) and Dixit (1980) focus on a single firm and potential entrant. The idea is that capacity serves as credible threat to potential entrants and thus an effective tool for entry deterrence. The difference between the two is, while Dixit claim that in post-investment situation, the incumbent will produce at the same level as its expanded capacity, Spence noted that it does not necessarily have to be so. Investment will be carried out by single incumbent firm to expand capacity and in the event of no entry; capacity does not have to be fully utilized, resulting in observable excess capacity.
Other, somewhat more recent, publications such as one by Gilbert and Vives (1986) assume a more practical setting in their approach on examining excess capacity by taking a non-cooperative Cournot oligopoly perspective. This setting is drawn upon based on the consideration that having few firms holding dominant position is more common than the single incumbent firm setting in the previous models. This dominance usually persists for a notable amount of time, in which the industry concentration level remains high despite technological progress. Therefore, it is believed that using the oligopoly model is more realistic for examining the incentives for deterring entry. This is quite precise in the case of cinema business in the Netherlands where for the past few years multiplexes (especially the four biggest chain namely: Pathé, JT, Utopolis and Wolff) have been dominating the industry both in terms of revenue and capacity1.
Moreover, although most of the literature mentioned so far focused on how excess capacity is an anti-competition tool, contesting argument against this concept also exists. Lieberman (1987) questions whether excess capacity does indeed deter entry and under what circumstances, it is in the incumbent firms best interest to maintain the excess capacity as such. The paper aimed to evaluate the frequency, in which the widely known theory actually happens in practice. Lieberman’s paper focuses on industries with certain characteristics that are believed to be most relevant for excess capacity to be formidable if it does prove to be effective as an entry-deterring mechanism. Such characteristics include high fixed cost, sizeable economies of scale, and a relatively small number of producing firms. Indeed, if we put the subject of this thesis on alignment with these characteristics, the cinema industry in the Netherlands fits rather well as an industry as specified by Lieberman.
It is further hypothesized by Lieberman that the reason for a firm to form excess capacity may be strategic or non-strategic. The cinema industry in the Netherlands proves to be a plausible candidate for examining both strategic and non-strategic motivations on holding excess capacity especially because it has somewhat lumpy and irreversible capacity as well as notable variations for demand. Strategic being, excess capacity persists for the purpose of enabling existing firms to get rid of current competition and/or deter entry by creating credible threat, and non-strategic when otherwise. It then argues that in nature, excess capacity is non-strategic and concludes that according to the empirical evidence carried out, entry barrier excess capacity theory is not very common in practice. That being said, it does acknowledge that the finding does not disqualify entry deterrence as an explanation just that it might not be completely effective because it undermine market growth and demand-related effects. This brings us to the alternative explanation on why would a firm hold excess capacity.
The alternatives explanation on why firms hold excess capacity is no less crucial than why excess capacity might be there to deter entry. Time and time again it has been pointed out that empirical evidence on the latter is sparse and mostly inconclusive. This is completely understandable because of the difficulties in
obtaining relevant data, something that proved an ordeal in the course of this thesis.
Empirically, the most prominent effort to distinguish between the alternative explanations and entry-deterrence explanation is by Conlin and Kadiyali Texas lodging case, which test whether the propensity to invest in entry-deterring capacity is a function of market concentration and market presence. The test is carried out in the second specification model of the paper, where it incorporates the share of capacity affiliated with certain brand, share of market capacity, incumbent opens property and new entrants opens property. Even so, though the model finds that more concentrated markets have greater capacity and firms with larger incentives to deter entry, that of larger market share do have larger idle capacity, which is consistent with the entry-deterrence explanation, this result is far from conclusive. Thus, a closer look on the alternative explanations might the best next thing.
Alternative explanations for excess capacity that often come up are for example expected demand growth. Firms decide on holding excess capacity simply because it anticipates future demand. It may also be the case that firms hold idle capacity at present because it has previously observed high demand. Another possibility is demand variability, which makes excess capacity as an instrument to cope with volatile demand. It may also be the case that firms may have idle capacity when there is collusion. Of course collusion by itself is illegal and punishable, however, highly concentrated market enables tacit collusion among existing firms to higher extent when it yields excess capacity, after all symmetry among firms in industry can support collusion.
Another interesting explanation for excess capacity worth mentioning is the idea that excess capacity works as a commitment to promote entry. This turns the rationale almost sideways and is a rather eerie alternative compared to what has been mentioned so far. But Thomas von Ungern-Sternberg (1988) expressed the concept that firm’s decision to hold excess capacity might be more benign than generally perceived. Under the premises that large firms have economic
relationship with other smaller upstream and/or downstream companies, excess capacity serves as a commitment to promote entry for these smaller industries. The promotion effort is therefore not intended towards those in the same level of business. The main idea behind this argument is that, these relationship forms considerable part or even interdependency in the total activities of the main firm. Therefore, by expanding capacity, dominant firms allow themselves to present an effective promise to their upstream and/or downstream counterparts to take part in the industry by ensuring that sunk costs will be covered. This is convenient especially in the absence of long-run contract. What is more interesting in this theory is that in this case excess capacity leads to Pareto improvement because incumbent firm and its upstream/downstream counterparts become better off after capacity is enlarged. Naturally, it is pointed out that the motivation behind holding excess capacity is not altruistic. The prospect of upstream/downstream companies not committing and/or not being competitive enough are mainly why signaling with excess capacity is reasonable. In addition to that, based on the premises, there are also other arrays of ways to ensure efficiency in regards to upstream/downstream companies and its market structure. Possibilities mentioned by von Ungern-Sternberg that might be in lieu of excess capacity include vertical integration, long –run contracts, and reputation.
Most of theoretical literature has placed excess capacity in connection with entry deterring mechanism. However, most empirical publications are still inconclusive on proving the connection and thus prove that to get a better picture one possibly has to resort to alternatives explanations.
Data and Methodology
The data on this thesis is for cinema industry in the Netherlands, collected from Nederlandse Vereniging van Bioscoopexploitaten (NVB) and Nederlandse Vereniging van Filmdistributeurs (NVF).
NVB and NVF jointly publish annual reports (jaarverslag), which give account of the annual revenue on national basis, number of cinemas2, number of screens, and number of seats on national and provincial basis. The report also contains the average number of cinema visit per person per year both on nationally and provincially. Furthermore, by additional request the NVB provided a more detailed data about the location of each cinemas3 and individual capacities.
Having such data in hand, the initial idea in examining how excess capacity plays its role in the cinema industry is as so; from the annual report it is known that there are several big chains of multiplexes4 which dominated the market for the last couple of years, therefore it seems reasonable to incorporate this dominance in the analysis of market concentration and later in relation to excess capacity. After all multiplexes do, somewhat conveniently, embody the idea of excess capacity. In order to do so, I turn to split the market for cinema in the Netherlands into provincial markets. Under the consideration that the dominant firms in the market, namely Pathé, JT, Utopia and Wolff, are dominating the market for period 2011-2013 both in revenue, number of visitors and capacity (shown on table I).
2 Throughout the annual reports there are several definitions of cinemas. Mainly they are categories such as bioscoop, filmtheater, filmhuizen, reis-en-openlucht bioscoop. All being the member of NVB, for this thesis I decided to include all the cinemas on the category in the calculation performed in later parts of this thesis. 3 Details about cinemas are based upon booking-group categories, sorted in company basis (for example Pathé, Go Filmbooking, and Onafhankeljk/Independent - though being independent might still be in the same category as major cinema chain in terms of booking-group) and categorized as such: bioscoop, filmtheater, and filmtheater klein (small). For later calculations, all types of cinemas are included. In terms of ownership, which would matter in market concentration calculation later on, cinemas are sorted to company that runs it, not the booking group it belongs to.
4 A multiplex is defined to be movie theatres with multiple screens, but since almost all movie theatres operating in the Netherlands have more than one screen, firms operating comparatively large cinemas are referred to multiplexes hereafter.
TABLE I – SHARE OF REVENUE, VISITORS, AND SEATS Revenue (x€1.000) Share Revenue Visitors (x1000) Share Visitors Seats Share Seats 2011 Pathé Bioscopen 105.820 44,08% 13.228 43,43% 34.592 26,48% JT Bioscopen/Luxor Theatres 24.347 10,14% 2.816 9,25% 15.044 11,52% Wolff Bioscopen 14.309 5,96% 1.828 6,00% 9.842 7,53% Utopia Bioscopen 10.251 4,27% 1.285 4,22% 6.895 5,28% Total Others 85.310 35,54% 11.301 37,10% 64.268 49,19% Total 240.037 100,00% 30.458 100,00% 130.641 100,00% 2012 Pathé Bioscopen 110.315 45,10% 13.493 44,15% 36.309 27,16% JT Bioscopen/Luxor Theatres 23.764 9,72% 2.771 9,07% 16.110 12,05% Wolff Bioscopen 12.967 5,30% 1.669 5,46% 9.292 6,95% Utopia Bioscopen 9.629 3,94% 1.136 3,72% 6.840 5,12% Total Others 87.925 35,95% 11.491 37,60% 65.152 48,73% Total 244.599 100,00% 30.560 100,00% 133.703 100,00% 2013 Pathé Bioscopen 115.173 46,16% 13887 45,06% 36038 26,32% JT Bioscopen/Luxor Theatres 23.579 9,45% 2816 9,14% 18158 13,26% Wolff Bioscopen 12.512 5,01% 1560 5,06% 9292 6,79% Utopia Bioscopen 9.092 3,64% 1069 3,47% 6284 4,59% Total Others 89.152 35,73% 11486 37,27% 67146 49,04% Total 249.508 100,00% 30818 100,00% 136918 100,00%
Perhaps the most interesting thing one can observe in Table I is that in all cases except Pathé (and this includes the independents categorized as total others), the market size in revenue and share of visitors are always lower than the market size based on number of seats. In Pathé’s case the share of the size of the revenue and visitors almost doubled compared to its share of seats. Pathé is indeed the largest among the big chains, with 22 cinemas operating in the Netherlands compared to Wolff and Utopia with 9 and 5 respectively at the end of 2013 and though JT has 21 cinemas running, its cinema size is only roughly half of Pathé’s. In addition to that, the revenues shown in this table are only those of ticket sales, so other revenues coming from other activities that commonly associated with cinemas such as advertising, food and beverages sales, etc. are not incorporated as revenue components.
Furthermore, it seems to be the case that the big chain multiplexes operate more in some provinces than others, in fact there are provinces which cinemas are not associated with any of the four dominant firms. Therefore, having twelve markets in hand, the first step towards answering the question “how concentrated is the cinema industry in the Netherlands?” is calculating the Herfindahl-Hirschman Index (HHI). HHI is a measure of market concentration, which is calculated by summing up the square of firm’s market share (s; can be in terms of revenue, production or capacity) relative to the industry.
𝐻𝐻𝐻𝐻𝐻𝐻 = � 𝑠𝑠𝑖𝑖2 𝑁𝑁
𝑖𝑖=1
HHI is often used to assess competitive situation (notably by the European Commission for the guidelines of horizontal mergers).
The HHI in this case are calculated based on the number of seats available in each cinema in the Netherlands with Dutch provinces as the market definitions. When calculating HHI, different cinemas that belong to same company within the dominant firms are considered as one firm5. First thing that comes in mind regarding this method is probably why use the number of seats for HHI calculation. There are of course other approaches to measure market share for HHI calculation; from using other variables like revenue and production output, to a more sophisticated method like Lambin and Schuiling’s (2007) share development tree. However, because of the data availability consideration, using the capacity or more precisely number of seats per cinema seems to be the most appropriate approach for this occasion. After all, as stated by Elhauge and Geradin (2011), market share measurement should be based on the best available indicator.
5 From previous specification of cinemas in (3) all cinemas are included in the calculation. To focus more on the role of dominant firms, all cinemas belonging to the big four are accumulated into its respective firm and other cinemas (in this case the independents (onafhankelijk) are calculated as individual firms.
Moreover, since the data available for writing this thesis is quite limited (the primary data consists only of annual reports between 2011-2013 and list of individual cinemas dated in 2014), it is often the case that the calculation of the yearly HHI requires manual collection of data. This involves meandering secondary sources (such as local news websites and individual cinema websites6) for information of newly opened and newly closed cinemas, as well as cross checking the number of the cinemas in yearly reports, the dataset, and the news in order to get correct dates and thus more precise HHI.
The process proves a bit enervating because only changes in big chain multiplexes are documented in terms of overall number in the annual NVB reports. Even so, since markets are defined on provincial basis, manual search to identify cinemas and to which market it belong was still needed. Furthermore, due to the large amount of independent cinemas and limitation of data, manual collection for individual cinemas is rather impractical. Therefore, a t-test was performed to analyze whether the differences in number of seats for individual cinemas are significant each year (differences are insignificant at 5% level). Nevertheless, for provinces with less than 10 of cinemas operating and with recorded change in capacity according to the annual report, manual searches were still done. Overall, for the period 2011-2013, manual data collection found eight changes (opening and closing) with five belonging to big chain multiplexes.
Next, using the average visit per person per year we can calculate the idle capacity. Taking the number of population of each province and multiply it with the average visit per year, we get the total yearly visit. Then, available capacity is procured by multiplying the number of seats available to 365 (assuming cinemas open every day of the year). Idle capacity is then the difference between the two. Loosely, the idle capacity is the available capacity minus the utilized capacity, which is the annual visit per year. Then, putting it on the percentage terms idle capacity is divided with the available capacity. Summary statistics of the data is provided in Table II, which contains the annual information of the number of cinemas, the number of seats, percent of idle capacity and Herfindahl index.
TABLE II – DESCRIPTIVE STATISTICS BY YEAR
Year
Variables 2011 2012 2013
NL Number of cinemas Number of seats 130.641 256 133.703 256 137.631261 7
Idle Capacity (%) 36,13% 37,38% 38,30% Provincial HHI and excess capacity based on number of seats8 Noord-Holland 1611,0 1797,5 1797,5 22,77% 19,34% 19,72% Zuid-Holland 2839,7 2662,3 2662,3 16,24% 19,17% 18,95% Noord-Brabant 1163,1 1163,1 1163,1 36,58% 39,77% 39,82% Gelderland 1101,7 1266,4 1266,4 49,81% 46,62% 48,28% Utrecht 1860,3 1510,7 1510,7 32,83% 45,93% 42,41% Overijssel 1793,2 1793,2 1793,2 49,78% 48,92% 48,96% Limburg 1066,6 1066,6 1019,1 55,59% 56,14% 61,64% Groningen 1936,2 1936,2 1936,2 49,23% 52,11% 53,13% Friesland 2133,2 2133,2 1973,7 48,22% 54,29% 54,75% Flevoland 4662,3 4662,3 4184,0 52,80% 57,13% 60,86% Drenthe 2810,1 2810,1 2810,1 54,82% 56,91% 57,11% Zeeland 2718,8 2718,8 2609,3 55,57% 56,44% 56,33%
Originally, it is the intention of this thesis to run further regression similar to the first specification model of Conlin and Kadiyali paper and add a dummy variable, which incorporates the existence of big chain multiplexes. The desired regression would be as such:
(% Idle capacity) p,t = α + β1 (HHI) p,t + β2..n (X)p,t + βn+1(M)+ ε
The dependent variable is market at province (p’s) percentage of idle capacity at time t. Then market at province p’s Herfindahl index based on capacity (seats) in
7 One will notice discrepancy between the number of seats between Table I and Table II; this inconsistency is statistically insignificant at 5% level and appeared because it comes from two different source
documents. Both documents are published and authorized by NVB; the difference being one (the one appear on Table II) is released on a later date (2014).
year t becomes the dependent variable that measures market concentration. If this model applies in accordance to popular theoretical literature, entry-deterrence motives are in play if the expected value of β1 is a positive number. Additionally, the model also includes vector X, which contains variables such as population, economic growth, unemployment rate, and per-capita income in province p at time t. These variables control for demand factors. Preferably, vector X also has cost factor elements to it. Variables, in this cinema case might be something in the line of business tax rates, construction cost, or hourly wage rate. Specifically in this model, a supplement variable regarding dominant players’ presence is added, dummy variable βn+1 measures 1 if there is one or more big chain multiplex present in the province and 0 otherwise.
This model in some likelihood has the potential to work had the data per cinema been available. Unfortunately, the precise per cinema data the industry proves to be unavailable to public. Therefore I decided to drop the variables in X and use the remaining variables to try to run the regression and as an approximation of the full analysis I present the result from somewhat derivative but more feasible approach of correlation analysis.
Limitations & Result
Before coming into the result, it is important to acknowledge the limitations that hinder this thesis from producing a more precise result. Naturally, the most influential constraint is the limited available data. This proves to be quite damaging, considering that the extensive regression could not be performed because of the absence of individual per cinema visit data and the simplified regression with small number of observations is bound to produce insignificant result in terms of standard errors.
Moreover, in regards to the methodology, there is an issue concerning the calculation of idle capacity, which multiplies the number of available seats to the number of days. This might be considered as somewhat arbitrary in the sense that it does not incorporate the possibilities that seats in cinemas are used several times during a day. However, since all cinemas are calculated in this manner, this is not particularly detrimental in determining the direction of the effect. Nevertheless, this approximation might have been improved had the data been more comprehensive.
Additionally, in terms of market, in this thesis I split the market for cinema in the Netherlands according to provincial basis. This inadvertently put each province at an even basis in terms of cinema distribution. As it turns out, this is not the case because the distribution of cinemas in the Netherlands on provincial term is bumpy (for example: North Holland has over 50 cinemas while Flevoland only has 5). Bearing in mind that initially the intention to use provincial markets is to emphasis on the effect of dominant player in the market, I did use the frequency of visit on provincial basis. This might to some extent capture the difference and even out the prominence of each province.
Lastly, in relation to running the data, under the simple regression different observations are considered to be independent of one another. However, some observations in this analysis are undoubtedly dependent. Yearly progressions on each province are inherently connected and therefore independent assumption
is violated. Ideally, using a panel data analysis would be more desirable but for this instance it does not seem practical because of the small data set in hand. Thus, it is important to note that the p-value in the following result is not reliable because it brushed aside the dependency between observations.
Having the limitations laid out, Table III shows the result of the regression for idle capacity with HHI and the dummy variables, which indicates the presence of big chain multiplexes in the corresponding province as the independent variables.
TABLE III – Result of Regression on Idle Capacity
According to Table III, the estimates coefficient for HHI though extremely small is positive and statistically significant. Meanwhile the coefficient for dummy variable shows a negative figure in spite of having high p-value that makes it statistically insignificant. In terms of marginal effect, having big chain multiplexes operating in the province actually decrease level the excess capacity.
Following the regression, table IV, V, and VI show the correlation coefficient of idle capacity with HHI, idle capacity with dummy, and HHI with dummy variable respectively9.
9 Correlation coefficient of idle capacity and HHI is on yearly basis, however, in the dummy case, since the dummy variable results do not change in the course of three years, the coefficient is in overall terms.
Table IV Correlation Idle Capacity and HHI
2011 0,1120
2012 0,1936
2013 0,1636
Table V
Correlation Idle Capacity and Dummy Variable
Idle Cap. Dummy
Idle Cap. 1,0000
Dummy -0,2936 1,0000
Table VI
Correlation HHI and Dummy Variable
HHI Dummy
HHI 1,0000
Dummy -0,1319 1,0000
Interestingly, though significantly larger, the correlation results show a similar result to the result shown in the regression in terms of direction (negative or positive). The coefficient for HHI shows that additional points on HHI bring a positive effect towards idle capacity (increase idle capacity) and the dummy coefficient negative effect or having a big chain multiplex in the province decrease the idle capacity. What is more interesting is how HHI and presence of big chain multiplexes, have different effect on the idle capacity.
At first, one may by default associate having big chain multiplexes operating in the market with high market concentration but that is not entirely the case for the market of cinema in the Netherlands. For instance, it may be the case in the province of Flevoland that big chain multiplex equals high concentration. Flevoland has got high concentration with 4148,0 HHI in 2013 and a big chain multiplex running in the area. However, there are also provinces like Limburg with 1019,1 HHI in 2013 and big chain multiplexes in it. As an overall effect, the
correlation result shows that the presence of multiplexes is negatively related to the market concentration.
To come in terms with this, we might return to the limitation explanation about how in the market specification all provinces are hold equally, which is not the case. In addition to that, in some market (like Flevoland, Drenthe and Groningen), big chain multiplexes are comparatively larger than other cinemas, while in other markets (for example: North-Holland, North-Brabant and Limburg) they are not. This fact may also contribute to the result.
Furthermore, provided the limitations as mentioned, this result is not entirely unexpected. Considering earlier data shown in table I, the case of Pathé, which shares of revenue is notably higher than its market share in terms of seats shows a little preview of how efficiency manifest itself in big chain multiplexes. Being the largest in player in the market it is not impossible that the Pathé’s idiosyncratic feature resonance itself to the overall effect of big chain multiplex presence. That being said, it is also possible that people just associate themselves more with big chain multiplexes and prefer watching films there thus making capacity in big chain multiplexes more utilized.
Discussion
It becomes increasingly apparent during the course of this thesis that the statistical result shown would not be particularly satisfactory and statistically precise. This is not a surprise following the footsteps of previous studies that tried to empirically examine the market concentration, excess capacity and the reasons behind it. Nevertheless, running the regression and generating the correlation analysis showed an idea about how market concentration affects idle capacity, which brings us to the next question, “What is the reasonable explanation(s) to hold excess capacity in the cinema market in the Netherlands?”
Seeing the result so far, in relation to big chain multiplexes, entry-deterring capacity does not seem to fit the bill as the major explanation. The presence of big chain multiplexes, which essentially are firms that have reasons to deter entry actually shows negative correlation with the excess capacity itself, therefore it is sensible to venture out to other alternatives to explain the motivations for holding excess capacity.
First, we can look at the expected demand growth explanations. Seeing the progression of cinema visit in the Netherlands, it seems that this is quite a reasonable explanation. For the last nine years, the cinema visit has been constantly increasing10.
Graph I – Cinema Total Revenue and Visits
10 Data for visit and revenue from 2005 – 2010 are obtained also from NVB through its jaarboek
17000 19000 21000 23000 25000 27000 29000 31000 33000 0 50000 100000 150000 200000 250000 2005 2006 2007 2008 2009 2010 2011 2012 2013 Revenue (x€1000) Visits (x1000)
Second, we could also explore the volatility of demand. Excess capacity may be there to fend cinemas from ever changing demand. Escobari and Lee (2014) pointed out that demand uncertainty is the key source of excess capacity. It further noted that often in the case of industries with high volatile demand, high cost fixed capacity and excess capacity which becomes invalid once selling period is done11; sellers have to decide on capacity before demand is realized and ended up with unsold inventories.
If we see the markets in the Netherlands, the demands are quite volatile for different times of the year. The general trends as shown in graph II display that the number of visits spike up during mid-year and end of the year. This might be explainable by the holiday seasons. After all, according to the yearly research by Stichting Filmonderzoek12, people aged 18-23 (despite being demographically small in the country) hold the largest share for cinema market. However, Bell and Campa (1997) study on chemical processing industry suggested that volatility might have positive or negative effect on capacity investment (depending on assumptions like firm’s risk attitude). This puts the volatility explanation hangs in the balance. Additionally, it is also may be argued that cinema industry could even out the volatility of demand through pricing so that extensive capacity becomes less called for.
Graph II – Cinema Visits by 4-week Period
11 To illustrate the third characteristics in context of cinemas, assume that excess capacity (vacant seats) becomes no longer valid after the selling period (sales of tickets) ends or as a movie starts.
12 Stichting Filmonderzoek conducts research commissioned by NVF annually.
1000 1500 2000 2500 3000 3500 4000 2011 (x1000) 2012 (x1000) 2013 (x1000)
Third, particularly in the cinema industry, the relationship between the players in the market and upstream/downstream companies may also be interesting to look at. Following an idea mentioned by von Ungern-Sternberg (1988) suggests that excess the capacity is used to signal the upstream and/or downstream companies to enter. The most likely candidate for upstream companies is the film distributor. If the situation is as so, then the dominant player in the cinema market would have more seats to attract more distributors into supplying more films. This is unlikely to be the case for this instance since major film distributors, which comprise over half of the market for cinemas in the Netherlands, are mostly supplying to big chain multiplexes anyway. Dominant players, namely the big chain multiplexes, do not need to signal their upstream counterparts.
Nonetheless, if we use the idea in reverse, this might prove to be quite a sensible rationale. Independent cinemas account for almost half of the market share in terms of seats but as seen in table I, it does not have that big of a share when it comes to revenue. Undoubtedly, independent cinemas, especially smaller ones do not have such strong relationship with major distributors as the dominant players and having excess capacity might signal in favor of getting more films supplied. Having said that, of course there is also the matter of the distribution cost and the cost of running cinemas in general, which according to de Roos and McKenzie (2014) is complicated because it depends on the type of movies played and the timeframe of which it runs. This might be the reason why major distributors do not supplied extensively to smaller cinemas in the first place.
Conclusion
This thesis aimed to examine Dutch cinema market concentration, excess capacity and the motivations for holding it especially in regards to entry deterrence. It is apparent that most existing theoretical literature point in favor of entry deterring motivation for holding excess/idle capacity although empirical studies has yet to give a definite answer to disqualify other alternatives. Having that in mind, this thesis then embark on inspecting the relationship between market concentration as indicated by Herfindahl Index, idle capacity, and the existence of big chain multiplexes in the market for cinema in the Netherlands on basis of provincial market.
The tasks proves to be a tricky one because of the limitation of data available, nevertheless; after running a simplified regression and constructed correlation coefficients between the independent variable and idle capacity, the result shows that market concentration has positive effect on idle capacity. Meanwhile the coefficient variable, which indicates the presence of dominant player in the market, shows that having a big chain multiplexes in the market reduces idle capacity. This result is quite unique because it shows that in the Netherlands, excess capacity in cinemas turn out to not be quite driven by entry deterring motivation.
What followed were the alternative explanations to fill in the gap for motivation and that seem to be more in context for the Dutch cinema market. These alternative explanations include the constant growth demand for cinema, demand volatility as well as signaling the upstream counterparts.
Finally, the Dutch market for cinema proves to be quite impartial and does not show any misconduct or specific attempt to deter entry. Quite the contrary, the existence of dominant firms actually indicates better capacity utilization. Consequently, at the present moment policy intervention is not called for.
Bibliography
Bell, G.K., Campa, J.M. (1997). Irreversible Investments and Volatile Markets: A Study of the Chemical Processing Industry. The Review of Economics
and Statistics, 79 (1), 79-87.
Bram. (2012, 7 November). Pathé Amersfoort Officieel Geopend. Film
Totaal. Retrieved from http://www.filmtotaal.nl/artikel.php?id=28160 Cinema Middelburg. (2014). Cinema Middelburg: Info. Retrieved from
http://www.cinemamiddelburg.nl/info
Conlin, M., Kadiyali, V. (2006). Entry-Deterring Capacity in the Texas Lodging Industry. Journal of Economics and Management Strategy, 15(1), 167-185.
De Roos, N., McKenzie, J. (2014). Cheap Tuesdays and the Demand for Cinema. International Journal of Industrial Organization, 33, 93-109.
Dixit, A. (1980). The Role of Investment in Entry-Deterrence. The
Economic Journal, 90(357), 95-106.
Door, H.B. (2011, 27 October). Wolff Bevestigt: Grootste Bios. Bouwput
Utrecht. Retrieved from http://www.bouwpututrecht.nl/bu/2011/10/27/wolff-bevestigt-grootste-bios/
Elhauge, E., Geradin, D. (2011). Global Competition Law and Economics. Oxford: Hart Publishing.
Escobari, D., Lee, J. (2013). Demand Uncertainty and Capacity Utilization in Arilines. Empirical Economics, vol. 47(1), 1-19.
Gilbert, R., Vives, X. (1986). Entry Deterrence and the Free Rider Problem.
The Review of Economic Studies, 53(1), 71-83.
JT. (2014). Over JT. Retrieved from https://www.jt.nl/over-jt
Lambin, J.J., Schuling, I. (2007). Market-driven Management: Strategic and Operational Marketing. London: Palgrave MacMillan.
Lieberman, M.B. (1987). Excess Capacity as a Barrier to Entry: An Empirical Appraisal. The Journal of Industrial Economics, 35(4), 607-627.
Nederlandse Vereniging van Bioscoopexploitaten, Nederlandse
Vereniging van Filmdistributeurs. (2011). Jaarverslag 2011. Retrieved January 8, 2015, from: http://www.filmdistributeurs.nl/files/574pokj5.pdf
Nederlandse Vereniging van Bioscoopexploitaten, Nederlandse
Vereniging van Filmdistributeurs. (2012). Jaarverslag 2012. Retrieved January 8, 2015, from: http://www.filmdistributeurs.nl/files/1652jmlml.pdf
Nederlandse Vereniging van Bioscoopexploitaten, Nederlandse
Vereniging van Filmdistributeurs. (2013). Jaarverslag 2013. Retrieved January 8, 2015, from: http://www.filmdistributeurs.nl/files/1530qycw23.pdf
Nederlandse Vereniging van Bioscoopexploitaten. (2013). Bioscopen
Nederland (leden NVB) op Boekingsgroep. Retrieved December 29, 2014, from:
http://www.nvbinfocentrum.nl/uploads/files/201402_stoelen_&_doeken_per_b oekingsgroep.pdf
Omroep Flevoland. (2012, 19 September). Utopolis zegt huur op van bioscooppand. Retrieved
from http://www.omroepflevoland.nl/Nieuws/94752/lelystad-utopolis-zegt-huur-op-van-bioscooppand
Slieker Film. (2013, 22 March). Slieker Film is Geopend! Retrieved from http://sliekerfilm.nl/nieuws/slieker-film-is-geopend/
Spence, A.M. (1977). Entry, Capacity, Investment and Oligopolistic Pricing.
The Bell Journal of Economics, 8(2), 534-544.
Stichting Filmonderzoek. (2013). Bioscoopmonitor 2012/2013: Een
Onderzoek Opdracht van de Nederlandse Vereniging van Filmdistributeurs en de Nederlandse Vereniging van Bioscoopexploitaten. Retrieved January 20, 2015,
from http://www.filmdistributeurs.nl/files/561bp5udu.pdf
Ufkes Apeldoorn. (2012, 25 May). Opening bioscoop CineMeerse Hoofddorp. Retrieved from
http://www.ufkesapeldoorn.nl/nl/nieuws/detail/opening-bioscoop-cinemeerse-hoofddorp
Von Ungern-Sternberg, T. (1988). Excess Capacity as a Commitment to Promote Entry. The Journal of Economics, 37(2), 113-122.
