Global Cinema Distribution: A Large-Scale Data-Driven Study of Film Screenings

Deakin University – School of Communication & Creative Arts – Melbourne, Australia

Doctoral degree awarded with no revisions on May 23, 2019

The use of big data to investigate the contemporary film industry is an area of recent academic interest. This interest has led to the emergence of new interdisciplinary studies, such as this project, combining the fields of digital humanities, new cinema history, and cultural economics. More specifically, newly available film distribution datasets have opened innovative avenues for research into the global flow of cinema.


This thesis establishes a method for investigating and analysing the temporal and spatial nature of contemporary film distribution at three scales: global, national, and international (US–Australia). It uses a large dataset of cinema showtimes that registers all cinema screenings in 48 countries spanning over 2.5 years from December 2012 to May 2015. This thesis engages with these datasets using innovative visualisation and computational techniques.


This thesis explores the movement and exchange of diverse contemporary cinema content, including often overlooked movies with low earning power and those produced outside the US. In this thesis, a feature film’s distribution is defined using four measures: the number of screenings, the length of the theatrical run, geographical spread, and release delay. Movie, distribution, and production factors that shape a film’s circulation, such as running time, genre, the size of the distribution company, the season of release, and production origin are also included in the study. 


A global sample of 3,424 films released in 2013 is first studied to reveal worldwide patterns of circulation. Five international distribution strategies are identified through cluster analysis: “short limited release,” “medium limited release,” “long limited release,” “wide release,” and “blockbuster.” 


A national sample comprising of 18,142 movie–country pairs is then examined to compare cinema distribution across 40 markets. Key factors that help explain exposure in any country, and in each market, are identified through multiple regression analysis. The length of run, geographical spread, and release delay are found to have a large effect in most countries, while the impact of movie, distribution, and origin characteristics varies across markets.


Finally, a sub-sample of 231 American films that screened in the US and Australia—including box office information—is analysed to model the trade relationship between the two markets, differentiating between the behaviours of high and low earning movies. Large cultural differences are detected between American and Australian audiences using non-parametric statistics. A relationship between exposure and box office is also proven by multiple regression analysis to be strong within both countries, although only moderately present for high earning films between the countries. The effects of movie, distribution, and origin characteristics are found to differ across markets and box office segments. 
 

This thesis demonstrates the benefits and insights gained from analysing the diffusion of movies around the world in both temporal and spatial terms. It uses film-related data that is usually reserved for written communication to be visually analysed, interpreted, and displayed in a way not previously explored. It has been found that by analysing multivariate cinema information it is possible to produce new insights into the geographic and temporal patterns and relationships present in the data. This thesis shows how engagement with quantitative data from different viewpoints can yield new insights about global cinema distribution. Findings from this study have the potential to inform cinema industry players on several issues including the optimal distribution strategies for certain types of movies, and the best release tactics for certain countries.

Keywords: cinema distribution, global film industry, movie business, international circulation, film screenings, big data, data visualisation, multiple regression

Supervisors:

  • Prof Deb Verhoeven (Canada 150 Research Chair in Gender & Cultural Informatics, University of Alberta)

  • Dr Bronwyn Coate (Senior Lecturer in Economics, RMIT University)

  • Dr Victoria Duckett (Senior Lecturer in Screen and Design, Deakin University)

Examiners:

  • Prof Alan Collins (Head of Nottingham Trent University Business School)

  • Prof Jaap Boter (Associate Professor in Marketing, Vrije Universiteit Amsterdam)

- Prof Alan Collins

This is among the best and most readable PhD theses I have examined.

This is indeed a very decent thesis, based on an impressive dataset.

- Prof Jaap Boter

- Prof Alan Collins

Most of my immediate critical responses to the work relate to admiration and awe of the ambition and coverage of the thesis.

- Prof Alan Collins

The work set out in this thesis is explicitly interdisciplinary and accurately described by the author as situated at the nexus of digital humanities, new cinema history and cultural economics.

- Prof Alan Collins

This work adds value conceptually to the cognate extant literature, as well as adding value by virtue of the novel empirical engagement (and sheer depth of engagement) with the combined data from Kinomatics and IMDb.

- Prof Jaap Boter

The thesis is not only detailed in its data and descriptions, but also in the several theoretical streams it positions itself against and a lot of references are extensively discussed and compared.

hello@vejune-zemaityte.com​

© 2020 by Dr Vejune Zemaityte