Methodological introduction



1 - Introductory remarks
Differently from the forewords, this introduction is of a more technical nature. Not only in the idiomatic usage in the film industry, but also in statistical publications different definitions of the same concept (e.g. “multiplex” and “co-production”) are used, which cause confusion and misunderstandings. With the “European Cinema Yearbook” we also want to contribute to the unity of a reasoned terminology. We are aware of the fact that once different concepts are placed in relation to one another, they appear to be more complicated than at first sight (e.g. the significance of an international comparison of average admission prices and of the meaning of “screen density”). However, for those who use this statistical Yearbook for more than a number of crude statistical figures, a more analytical approach is indispensable. Therefore, we hope that the “yellow pages” of the “European Cinema Yearbook” are appreciated as a contribution to a better understanding of developments in the cinema industry.

2 - The contents of the Yearbook
The 2008 edition consists of a series of comparative tables, covering the period from 1989 up to and including 2007, which offer a clear and immediate picture of the overall situation and of the trends in the cinema industry in 34 countries of Western, Central and Eastern Europe and of the Mediterranean Rim. Following a long tradition, this edition is completed by a section dedicated especially to multiplex cinemas, documenting the situation at 31 October 2007 and at 1° Jenuary 2008. A section devoted to digital cinema worldwide also appears, listing the digital cinemas equipped with DLP CinemaTM technology as at 30 June 2008. This introduction deals with a number of statistical problems, some of which are controversial. The purpose of this Yearbook is not only to give reliable data, but also to clarify concepts that are frequently the object of misunderstandings and mistaken interpretations. In the statistical part many notes are included, many of which indicate restrictions, making some figures less significant, though more realistic. Some notes, however, are of an essential nature, and refer to sections of this introduction. As is customary in statistics, data that is not available is marked by a dot (.); figures of exactly zero are indicated by a dash (-), and those approaching zero by 0, 0,0 or 0,00. A description of the working definitions of the terms used in this Yearbook follows this introduction.

3 - The use of sources
To avoid inaccuracy, we gather our data as closely as possible to primary sources, starting out from the national and international associations of the European cinema industry. Since, in many cases, they are members of MEDIA Salles, this facilitates our work together. MEDIA Salles is also in contact with other national and international bodies. In cases where the same kind of data is supplied by more than one organization, one of which being a national exhibitors’ or distributors’ association, we consider one of the latter two as being the most reliable source of data concerning cinema-going, unless the figures provided by it are clearly questionable. Thus, by relying on the collegiality within the cinema industry, the Yearbook is, as it were, a collective publication of the respective associations, and in this way information is gathered that would not have been obtained otherwise. However, since some associations and institutions do not have this range of data available and in some countries one seems to place more value on statistical information than in other countries (i.e. is reluctant to publish it), we have in certain cases been obliged to include data provided by other agencies or taken from publications, which must be considered secondary sources, if not tertiary. On the other hand, firms providing data concerning technical equipment, advertising receipts etc., are the primary sources in their respective fields. The consequence of this view is that in this framework a governmental statistical agency is not to be seen as a primary source just because it is governmental, but only if it is directly responsible for the content of the information (as is the case with CNC in France and FFA in Germany). Furthermore, we observe that in some publications the use of sources is not always as clear as it should be and leads to questions as to how the data quoted has been obtained. In principle, the practice of indicating oneself as the source without mentioning how data has been obtained (by calculation, estimation or even only by subjective evaluation) is judged by us to be inappropriate. Another practice we are critical of is indicating that data is not available whilst in fact it is. This practice, which is to be found in some statistical publications, unjustly throws a bad light not only on the publication concerned but also on the primary sources involved (e.g. an exhibitors’ association). Nevertheless, when information from primary sources was lacking, we made use of such secondary ones, leaving it to the reader to evaluate the reliability of the data concerned. Approximate data is indicated by the prefix c. and it is written in italics. For a number of countries, the figures for the gross distribution revenue and/or the average film rental are rough estimations. In these cases (indicated by c.) we chose the figure judged by us to be the most reliable. Nevertheless, these figures are not to be seen as very significant. Lastly, we should point out that, whilst until 2001 the sources used for Germany were FFA as well as SPIO, with the consent of the latter we now use mainly the FFA. As regards Switzerland , Procinema figures were used until 2003, whilst since 2004 we have been using as our source the Federal Office of Statistics (OFS), which also provides information on Liechtenstein.

4 – Currencies
In the comparative tables, data expressed in money values – up to 1998 – is shown in ECU and in EURO from 1999, whilst in the national reports it is given in the respective national currencies (for the Euro area, up to 2001). Due to changes in these currencies’ ECU/EURO exchange rates, it is, therefore, possible that the fluctuations of certain data (e. g. average admission prices) are not exactly equal to those shown in the respective national currencies. To get an impression of the development in ticket prices and box offices in one country, one must look at the chapters about individual countries (“European Cinema Yearbook”, “Source Document” –, whilst the comparative tables serve in the first place to compare the figures in the different countries. Relative ECU/EURO exchange rates were taken for the last month of the year in question (see currency table). The figures for which secondary data sources in US dollars were used are not shown in these tables, since the relevant US dollar exchange rates used in their calculations are unknown and therefore cannot be accurately converted into ECU/EURO equivalents. The use of ECU (and of EURO since 1999) for the comparative tables seems obvious because this Yearbook is about European data. Furthermore, the use of US dollars in comparative tables, as found in some other publications, is less appropriate for comparisons between data concerning successive years because the exchange rate of the dollar has fluctuated much more than that of the ECU.

5 - Art cinemas
It also our custom to provide an entry on art cinemas for each country. The term “art cinema” is however hard to define. The CICAE (International Confederation of Art Cinemas in Europe ) does not have members in all countries, so we cannot use membership of CICAE as a standard. We therefore chose to include in this category those cinemas which are designated by the associations to which they belong as “art cinemas”. Another problem in this area is that some art cinemas are non-commercial and run on government subsidies, whilst others are commercial and sometimes receive a partial subsidy. Unless otherwise indicated, we have only included commercial “art cinemas”, i.e. those which are run by enterprises, whether partially subsidised or not.

6 - Market shares of European films
Particular attention has been given to market share figures for national and non-domestic European films. However, as these figures do not give an impression of the effort which was made to show European films (looking at the market share figures of US films in certain countries one might think there are almost no European releases in these countries), we included, in national profiles, whenever available, information on the number of playing weeks or on screenings of European films, as well as a table showing a breakdown of admissions and box office of European films, based on country of origin. The aim of this scheme is to show the circulation of European films outside the country of origin in Western, Central and Eastern Europe . It should be kept in mind that the market shares of domestic films and of non domestic European films are arithmetically also dependent on the success of US films. Taking this into account, the results of European films in European countries are more stable than is indicated by their market shares1. Whilst in the vast majority of cases it is not difficult to decide if a film is to be regarded as European, the decision as to the countries in which a co-production should be listed as domestic, is often arbitrary. A complicating factor is that in film statistics market shares of domestic films are indicated inclusive of coproductions. This is no problem in the country from which the main contributions to the production of a film originate and where the film also reflects something of that country’s culture. However in a number of cases a film is also listed as a co-production in countries from which the contribution is marginal or does not concern content (e.g. purely financial), thus causing it to remain foreign in the public’s eyes. This means that also in such cases the admissions to a co-production are counted as domestic in more than one country, thereby favouring the market shares of domestic films, whilst making the market shares of non-domestic European films (which get much political attention) appear lower than they really are. Because the number of co-productions is increasing, there is a growing difference between the statistical registration of the market shares of non-domestic European films and what would be registered if a correct operational definition of co-production were used. Therefore we recommend that the European Commission should at least introduce a guideline on this matter2. A correct registration of market shares of European films is also impeded by counting all co-productions of one or more European producer(s) with a US one as European films, whilst, in fact, part of these films have a purely American character.

7 - Admission prices
Converted to EURO the average admission prices vary considerably from country to country. However, the differences are much smaller when the prices are adjusted to Purchasing Power Parity. This variation is again different when the prices are adjusted to GDP per capita, which is another way of comparing the average “real” ticket prices. These calculations demonstrate that the significance of differences between the average ticket prices depends very much on the way these prices are compared. A further comparative table regards annual per capita spending on cinema tickets. Additionally, in the chapters on certain countries we have added the highest ticket price charged, usually in the capital cities.

8 - Screen density
Screen density is usually indicated as the number of inhabitants per screen (being its reverse). As is rightly stated already in Screen Digest of September 1994 (p. 202) “the relationship between screen density and ticket sales is especially pertinent at a time when multiplex operators are implementing expansion plans in many parts of the world”. At first sight it appears that there is a direct link between screen density and admissions per capita, “but such a simple analysis would be to ignore some extremely important social and economic factors” (ibid., p. 203). One of these factors is undoubtedly the population density: in a densely populated area (especially if evenly spread), people have the choice of a number of cinema screens within a short distance. However, in countries (or regions) with a low population density, the screen density may be high if one uses the number of inhabitants per screen as a measurement, as is e.g. the case in Sweden and Norway . Between population density and screen density a considerable correlation ( 0.82 in 1991 and 0.84 in 1995)3 has been ascertained. From this insight, the following hypothesis can be deduced: “the higher the population density, the lower the screen density can be in order to bring about under otherwise equal circumstances (cet. par.) a certain number of admissions per capita”. The impact of the population spread has been studied more in detail in “Cinemagoing Europe” (Dodona Research 1994). The data given for Norway is a clear illustration: it is the third most densely screened country in Western Europe but Oslo is more densely screened than the national average for all but four European countries (pages 17-18). However, because even insiders often use rather superficial reasoning about the relation between the number of people per screen and admissions per capita, which can have severe consequences for the whole industry, since the third edition of the Yearbook another way of measuring screen density has been introduced, namely the density of screens per square kilometre. As to the number of screens in a country, Dodona Research, 1994, remarks that the absence of data on full-time and part-time screens is critical (p. 17). Of course, not only should closed cinemas not be counted, but it is also as well to exclude cinemas giving very few screenings. However, even full-time screens form a rather heterogeneous group as regards the number of performances given. For instance a screen with only one performance every day of the year is more similar to a so-called part-time screen with three daily performances during six months than to a screen with continuous performances. Therefore, a mere division between full-time screens on the one hand and part-time screens on the other, would not result in a statistical improvement. Besides, it is very difficult to obtain reliable data on the yearly number of performances in each country. This statistical imperfection is acceptable and in our opinion does not constitute a major problem. However, as far as this was published, since the 1997 edition we have published this type of information.

9 - Advertising receipts
The data regarding these revenues is not fully comparable, because for some countries the exhibitors’ shares are given, whilst for other countries only the figures concerning total advertising revenues are known, sometimes even including the production costs of advertisements.

10 - Multiplexes/Megaplexes
Up to now there is still no definition of “multiplex” that is generally employed in all countries. This is regrettable because the terms “cinema complex”, “multiscreen”, “multiplex” and “megaplex” or “mega cinema” are used and often misused, causing misunderstanding among journalists and thus also among the public. A special term is only meaningful if it is to be used to make a significant distinction. Based on econometric research commissioned by Media Salles, London Economics came to the conclusion that “the multiplex effect kicks in only with 8 or more screens, not with 6 or 7” . (“Cinema Exhibition in Europe , White Book of the European Exhibition Industry”, MEDIA Salles 1994, second edition, Vol. 2, p. 48). A similar point of view is to be found in Dodona Research 1994 (“Cinemagoing Europe”, p. 3). The criteria to be added to the definition of “multiplex” or to define the term “megaplex” can, of course, only be more or less arbitrary. To encourage a more generally accepted use of terms, the General Assembly of the Union Internationale des Cinémas (UNIC) held in May 1998, decided unanimously that, based on the above mentioned research, a conventional cinema must have eight screens or more to be called a multiplex, and twice this number, thus 16 screens or more, to be called a megaplex. Of course, the General Assembly of UNIC was aware of the fact that more criteria are of significant importance, e.g. parking facilities, stadium seating, air conditioning, big screens, spacious foyers (see e.g. J. Ph. Wolff, “Of multiplexes and multiscreens”, UNIC, Paris, Dec. 1993 – also available in French). However, from a statistical point of view, qualitative criteria like this are very difficult to handle. Because there are not many sites with 8 or more screens that should not be called multiplexes because they do not meet those qualitative criteria, the category 8 or more is fairly homogeneous, which is a practical advantage. Admittedly, a negative consequence of this operational definition is that a cinema with less than 8 screens becomes statistically a multiplex, by the backdoor as it were, when one or more screens are added to the premises up to a total of 8 or more screens. The reason for assigning a special term to multiplexes with a large number of screens is that they apparently have a bigger catchment area than smaller multiplexes. The figure 16 is not based on the result of scientific research, but is more or less arbitrary and could also have been 15 or 17. In the meantime the criterion of eight screens or more has been adopted in most European countries. In France it is (together with the condition of possessing more than thousand seats) even the official definition of the Centre National de la Cinématographie (CNC). This definition is also used by the Social and Cultural Council of the Dutch Government (May 2005). In the UK , however, the term “multiplex” is mostly used for every newlybuilt cinema with five or six screens. This is regrettable not only because no unity in the use of the term can be achieved like this at least in (Western) European countries, but also because the criterion of five or six screens marks an arbitrary difference without referring to a distinguishing feature. Furthermore, the criterion “purpose built” or “newly built” should not be used because only the features of a conversion, or enlargement, of a cinema should matter. (For deviant uses of the term “multiplex” see also the essay “Multiscreen, multiplex, megaplex?” introducing the chapter “Multiplexes in Europe ”). To give a global impression of the extent to which multiplexes have penetrated the respective cinema markets, the Research Group of MEDIA Salles introduced the concept degree of penetration of multiplexes, meaning the number of screens in multiplexes as the percentage of the total number of screens. These degrees are indicated in the table “Density of screens in multiplexes”. Nonetheless, it can be understood that this gives only a rough insight indeed, if one realises that changes in this so defined degree are also dependent on changes in the number of screens outside multiplexes. A new item in the comparative tables (“Multiplexes in Europe”) are the quotients of the number of admissions per screen in multiplexes and the number of admissions per screen in other cinemas. These figures are a kind of indicator, namely of the relative effectiveness of screens in multiplexes in a country. An advantage of this indicator is that it is not dependent on the absolute values of the numerators and the denumerators, but only on the relation between these two figures, making it possible to compare different markets in this respect (e.g. in 2002 the average number of admissions per screen in multiplexes in Spain was much lower than in the Netherlands but, due to the bigger difference with respect to the number of admissions per screen in other cinemas, the relative effectiveness was higher in Spain). One should keep in mind that this indicator is of little significance in a year with a relatively substantial overall addition of screens in multiplexes opened in the course of the year, especially in the last part of it (this was, for example, the case in the Netherlands in 2000 and 2002). Furthermore, calculating this indicator is not meaningful when the degree of penetration of multiplexes reaches very high values, say > 65%, because the composition of the other, remaining cinemas affects their comparability (e.g. when there are hardly screens with mainstream programming outside multiplexes in the big cities anymore).

Dr Joachim Ph. Wolff

1 J. Ph. Wolff, “The exhibition of European films revisited” (paper presented in the MEDIA Salles seminar during Cinema Expo International, Amsterdam , June 1999).
2 This recommendation was also made by Dr André Lange (Eur. Audiovisual Observatory, Strasbourg ). See E.J. Borsboom and J. Ph. Wolff (eds.), “Proceedings of the Seminar on Film Statistics on 26 June 2002 in Amsterdam ” (Research Foundation of the Neth. Cin. Fed., July 2002) See also: J. Ph. Wolff, “Non domestic European Films on the West European markets” (European Cinema Journal, May 2002).
3 J. Ph. Wolff, “Production is Key in the Film Industry”, Lelystad, 1998, p. 300. (His earlier calculation for 1991 is quoted by London Economics in Vol. 2 of MEDIA Salles’ “Cinema Exhibition in Europe . White Book of the European Exhibition Industry”, vol. II, p. 15).