Filmgoing and Film Popularity in Antwerp and Ghent, 1952

International statistics show that Belgians went more to the cinema than cinemagoers in neighbouring countries. The Cinema Belgica database offers researchers plenty of data for calculating and analysing cinema's popularity in the country. In this story, film historian and economist John Sedgwick outlines his robust POPSTAT methodology for investigating cinema's popularity in the cities of Ghent and Antwerp in 1952.

By John Sedgwick

Cinemagoers queuing in front of the Capitole cinema in Ghent in October 1948. The cinema is screening Victor Fleming's Joan of Arc (1948), starring Ingrid Bergman. The film ran for three weeks. (source: Rijksarchief Gent)

Advertisement for Joan of Arc (1948) in the local newspaper De Gentenaar (October 28th 1948)

7th in international rankings of cinema attendances

A UNESCO publication of 1952 ranks Belgium 7th in international rankings of cinema attendances per head of population with 14 visits per annum, just behind Italy and ahead of the German Federal Republic, France, and the Netherlands. Headed by Great Britain with 28 visits per head, Italy and Belgium lead the list of non-English speaking countries. This story focuses on cinema's popularity in two Belgian-Flemish cities in 1952, Ghent and Antwerp. This story is based on a methodology aiming at understanding cinema's popularity by using statistical data.

International cinema attendance data show the medium's popularity in Belgium in terms of the number of admissions per head. (source: UNESCO data, reproduced in Browning and Sorrell, 1954: 136)

Box-office revenues not made public

The study of cinema's popularity rests on empirical foundations. Without pre-judgement, the essential question is: what films did people go to watch? And what does this tell us about their preferences? These questions are problematic because industry statistics concerning the box-office revenues generated by films on release were not made public, or they are no longer available. The best systematic evidence about cinema's performance in the post-World War 2 era was published annually in industry yearbooks associated with the trade journals like Variety in the form of a list of the rental incomes of 'Top Grossing' films. Of course, Variety's list was specific to the North American market.

POPSTAT methodology

For this reason, any attempt on the part of researchers to maintain that Film X was more popular than Film Y in a particular locality at a moment in time requires a proxy method of measurement. In the absence of box-office data, POPSTAT was designed to fill this gap. However, the technique yields other benefits. Not only does it allow researchers to rank order films in terms of audience preferences, but in doing so lays open how films circulated within localities and the hierarchy of cinemas through which they moved.

The POPSTAT Index of Film Popularity is calculated for a population of films screened in a population of cinemas, particular to a locality. It is computed for each ith film in period t based on:

  1. the box-office potential of each jth cinema at which the ith film plays at time t, relative to other cinemas. This known as the cinema weight a;
  2. the billing status b of the ith film at the jth cinema at time t;
  3. the length of run l of the ith film at the jth cinema at time t.

The symbol sigma Σ is a summation instruction.

Cinemagoers queuing in front of socialist Cinema Vooruit in Ghent in 1958. (source: Amsab)

Price information is essential in estimating film popularity. Essentially, the argument is that cinemas differ in revenue potential—the product of the average price paid by customers and the number of tickets sold. Ranking them in terms of this potential allows the distinction to be made between the box-office potential of cinemas along a continuum from box-office rich to box-office poor. Accordingly, analysis that treats all cinemas and, or screenings, as equivalents will not capture this aspect of the film distribution process.

Consumers paying higher prices

The Antwerp cinema Vendôme shows Jacques Tati’s Les Vacances de Monsieur Hulot (1953), as well as the evergreen Gone with the Wind (1939). (source: Paul Corluy)

Does a question arise from this empirical observation concerning why film consumers were prepared to pay higher prices to watch films than they needed to? The answer is to be found in the desire on the part of audiences to (a) watch films earlier, rather than later, in their distribution cycle – what economists' term 'time preference'; and (b) watch films in better-accoutred environments. When both conditions are met, audiences will watch films in box-office rich cinemas.

Comparable intensity of cinemagoing with British cities

Evidence presented in Table 1 below represents information about the plentiful supply of cinemas and seating in Ghent and Antwerp. Indeed, the statistics presented in this table show a comparable intensity of cinema and seating concentration with that recorded for British cities by Browning and Sorrell (1954) in their monumental statistical survey of cinemagoing in Great Britain at the time.

Table 1. The supply of cinemas and seating in Antwerp and Ghent in 1952
  Antwerp Ghent
City Population 308.079 218.232
Total Cinemas 52 38
Population per cinema 5.925 5.743
Total Seats 42.787 23.310
Population per seat 7,2 9,4
Sample Cinemas 30 33
Seats in sample cinemas 28.281 20.566
Seats in sample cinemas as a proportion of Total 0,66 0,88
Mean seats of sample cinemas 943 623
Median seats of sample cinemas 810 560
Sample cinemas max. seats 2.000 1.663
Sample cinemas min. seats 300 186
Sample cinemas with 1,000 or more seats 11 2
Sample cinemas with 500 or fewer seats 4 10
Not included in the sample
Non-sample cinemas (excluded) 22 5
Non-sample cinema seats (excluded) 14.506 2.744
Non-sample mean cinema seats (excluded) 659 549

Source: Cinema Belgica dataset
Notes: Population data are taken from census returns of 1947. Antwerp population and cinema totals include the suburb of Borgerhout. Ghent population and cinema totals include the suburbs of Gentbrugge, Ledeberg, and Sint-Amandsberg.

The Cinema Belgica dataset is drawn from daily newspaper listings in newspapers (De Gentenaar and Vooruit), from which we track the circulation of films. Unfortunately, this source does not include admission prices, without which it is not possible to calculate cinema weights. Another problem is that not all cinemas advertise their programmes in these papers.

Programming data from these newspapers is supplemented with information published by two national organizations engaged in cinema publicity and advertising—respectively, the Chambre Syndicale Belge de la Publicité Cinématographique and the Union Belge des Annonceurs. Their third annual report published in 1954 contains the national listing of cinemas broken into regional urban locations. Against each cinema entry, data about their seating capacity, number of weekly screenings, and quarterly attendance figures, expressed as weekly averages, is listed. Further, and most importantly for the purpose at hand, each cinema was attributed a classification code, which determined the price per meter that advertisers were required to pay for screened promotions.

Admission prices

This pricing information can be used to get around the absence of admission prices. Based upon the assumption that advertisers were prepared to pay more to have greater access to larger, high ticket price paying audiences, cinema advertising prices can be thought of as a proxy for the revenue potential of cinemas. Multivariate regression analysis is conducted to test the validity of this approach, in which cinema advertising prices are modelled as the dependent variable (Y), and seating capacity, weekly programming frequency, and weekly average attendances, as independent explanatory variables (X1, X2, and X3) – the product of all three can be understood as determinants of the box-office.

Cinema Weights

Unfortunately, the list of cinemas for which advertising price data is found in the Répertoire publicitaire du cinéma belge does not coincide fully with the list of cinemas for which programming data has been collected. Where this is the case, cinema weights have been awarded based on the type of films being screened at those cinemas, their vintage, and their position in the cinema hierarchy. These are set out below in Tables 2 and 3 and show ranges varying from 2.28 (Rex) to 0.33 (Rio) in Antwerp and Ghent 3.03 (Capitole) and 0.33 (Muide and Vriendenkring). The revenue capacity of cinemas varied considerably, by multiples of approximately seven and nine respectively in the two cities.

Table 2. Cinema weights and associated data for Antwerp cinemas for which programming data is available
        Weekly Attendances    
Cinema Cinema weight Seats Weekly screenings Q1 Q2 Q3 Q4 Mean weekly attendance Price per meter (fr.)
Ambassade 1,06 450 35 4531 4797 3309 5826 4616 65
Americain* 0,41 792             25
Anvers Palace 1,95 1494 35 9528 10095 8187 11975 9946 120
Astra 1,30 800 28 6391 5732 7807 6795 6681 80
Astrid 1,63 600 28 5816 7176 6843 6818 6663 100
Capitole 1,63 1148 49 8801 8368 7996 9266 8608 100
Century 0,81 790 16 3142 3167 3271 3886 3367 50
Coliseum 1,30 900 35 2276 1866 6675 5130 3987 80
Crosly* 0,41 700             25
Dixi 0,81 900 18 4464 4072 3912 4264 4178 50
Eden 1,06 820 35 4094 5645 3923 4825 4622 65
Empire 1,95 1567 35 9519 7858 7083 8959 8355 120
Festa 0,81 560 35 5419 4804 4479 5148 4963 50
Luxor 0,81 1000 17 3945 3214 2911 3302 3343 50
Majestic 0,65 1100 29 2790 2017 2915 2396 2530 40
Metro 2,28 2000 35 13130 13808 9197 13090 12306 140
National 0,65 920 18 3571 3290 3016 3970 3462 40
Odeon 1,63 700 42 7790 6805 6527 7223 7086 100
Pathé 1,63 1000 49 10675 8837 9790 12060 10341 100
Plaza 0,65 550 18 2364 2226 2575 2535 2425 40
Regina 0,65 300 35 2554 2754 2252 3260 2705 40
Rex 2,28 1065 49 13769 13499 12674 14168 13528 140
Rio** 0,33 500             20
Roma 1,30 2000 16 6836 7312 6603 7709 7115 80
Roxy 1,30 1825 28 6063 5567 5643 7486 6190 80
Rubens 1,63 1200   9562 6629 7620 7854 7916 100
Savoy 1,06 700 42 4227 4001 4379 5499 4527 65
St Jan* 0,41 700             25
Studio Movy 0,81 400 35 4229 4192 3753 5887 4515 50
Victory 0,65 800 17 1540 1410 1383 1754 1522 40

Sources: Cinema Belgica dataset; Répertoire publicitaire du cinéma belge, 1954
Notes: *Assume a category 1 cinema, charging 25 cents per meter for advertising/promotional films; ** assume a category 2 cinema, charging 20 cents per meter for advertising/promotional films.

Table 3. Cinema weights and associated data for Ghent cinemas for which programming data is available
        Weekly Attendances    
Cinema Cinema weight Seats Weekly screenings Q1 Q2 Q3 Q4 Mean weekly attendance Price per meter (fr.)
Agora 0,63 600 18 1571 1770 1335 1698 1594 25
Cameo 0,63 590 8 1790 1961 2020 2181 1988 25
Capitole 3,03 1700 28 10375 10892 8740 10965 10243 120
Casino 0,63 550 12 1431 1399 1311 1698 1460 25
Century 1,64 775 28 5817 4719 4992 5847 5344 65
City 1,01 530 8 3707 3573 3205 3742 3557 40
Eldorado 1,64 608 35 4454 4539 3001 4657 4163 65
Forum 0,63 540 8 1766 1702 1274 1872 1654 25
Ganda 0,51 550 18 2158 1954 1940 2319 2093 20
Ideal 0,63 491 8 2424 2361 2067 2709 2390 25
Leopold* 0,63 547             25
Lido 1,01 500 18 2873 3105 2513 3538 3007 40
Majestic 2,53 958 28 7588 7150 5940 7929 7152 100
Metropole 1,01 690 27 4783 4212 3798 4780 4393 40
Muide*** 0,33 186 8           13
Nord Palace 0,63 400 18 1703 1451 1535 1808 1624 25
Nova 0,63 550 5 1096 1152 968 1279 1124 25
Novy 1,01 560   4189 3737 3367 4321 3904 40
Odeon 0,63 420 10 1700 1579 1598 2138 1754 25
Pax 1,01 700   646 666 620 966 725 40
Plaza 1,64 750 28 3995 3522 3770 5002 4072 65
Rex 1,26 698 28 3724 3251 2948 4321 3561 50
Rio 0,63 460 8 1492 2004 1131 1235 1466 25
Ritz 0,63 374 12 637 776 578 1140 783 25
Roxy 0,63 420 11 1197 1404 1296 1502 1350 25
Royal 1,01 800 18 4853 4486 3775 5011 4531 40
Savoy 1,26 590 28 3949 3514 2785 4029 3569 50
Scala 1,01 850 18         4500 40
Scaldis** 0,51 332             20
Select 1,01 547 28 2732 2409 2525 3251 2729 40
Vooruit 2,02 1500 28 7389 7022 7726 9039 7794 80
Vox 0,63 450   1705 1588 1551 1882 1682 25
Vriendenkring 0,33 350         154 154 13

Sources: Cinema Belgica dataset; Répertoire publicitaire du cinéma belge, 1954
Notes: *Assume a category 1 cinema, charging 25 cents per meter for promotional films; ** assume a category 2 cinema, charging 20 cents per meter for promotional films; *** assume a category 3 cinema, charging 13 cents per meter for promotional films.

The next task is to apply them to the film programming data. For each ith film, a cinema weight is given for each jth billing, which, when multiplied by the length of run and billing status, gives rise to a POPSTAT value. A POPSTAT Index value is formed when POPSTAT values are aggregated across all jth cinemas at which each ith film appears. From this, it is possible to construct a popularity index, which can be compared to other popularity indices.

Further Research

This story suggests a method for investigating cinema's popularity in two Belgian cities in 1952. It can be applied to other cities and accordingly serve as a basis for doing comparative analysis. For instance, the divide in the diet of films consumed in Belgium during the German occupation can be investigated using these methods. POPSTAT also facilitates the exploration of historical changes in audience tastes and associated cultural issues. An example here could be the decline in the influence of religious and political silos, the classical “pillars” on cinemagoing habits.

Other Sources and Data

Researchers who are interested in the POPSTAT methodology can also look at:

  • European Cinema Audiences: the POPSTAT methodology is used in this comparative research project on seven cities across Europe, including Ghent;
  • Cinematek: the library of the Royal Film Archive or Cinematek contains dozens of Belgian trade journals and yearbooks where one can find statistical data on Belgium's film exhibition scene.

Further Reading

  • Browning, H., and Sorrell, A., (1954). 'Cinema and cinema going in Great Britain', Journal of the Royal Statistical Society, 117, 133 165.
  • Pafort-Overduin, C., Sedgwick, J., and Van de Vijver, L. (2018). 'Identifying cinema cultures and audience preferences: a comparative analysis of audience choice and popularity in three medium-sized Northern European Cities in the Mid-1930s ', TMG Journal for Media History, 21(1), 102–118.
  • Sedgwick, J. (2000). Popular filmgoing in 1930s Britain: a choice of pleasures. Exeter: University of Exeter Press.
  • Sedgwick, J., (2002). 'Product Differentiation at the Movies: Hollywood, 1946-65', Journal of Economic History, 62, 676-704
  • Sedgwick, J. (2006) 'Cinemagoing in Portsmouth During the 1930s', Cinema Journal, 46(1), 52–85.
  • Sedgwick, J. (2020). From POPSTAT to RelPOP: A Methodological Journey in Investigating Comparative Film Popularity. TMG Journal for Media History, 23(1-2), 1–9. DOI: http://doi.org/10.18146/tmg.776
  • Sedgwick, J., Pafort-Overduin, C. and Boter, J. (2012) 'Explanations for the restrained development of the Dutch cinema market in the 1930s', Enterprise and Society, 13(3), 634-671.
  • UNESCO (1952), Basic Facts and Figures. Paris: UNESCO.

Author

John Sedgwick is an economic historian who researches film history. He is interested in the ontological characteristics of film as a commodity and film popularity and what these tell us about comparative audience preferences and industry processes and structure. He has written a monograph, edited an anthology, and published articles in leading academic journals.