Movies from foreign lands: Computing the geographies of cinema

By Andrea Ballatore (King’s College London) and Giovanni Quattrone (Middlesex University)

President Trump’s erratic protectionist drive has turned toward “movies coming into our country that are produced in foreign lands”, with proposed tariffs aimed at non-American productions (The Guardian, 2025). The administration’s goal appears twofold: bolstering domestic film production and, unsubtly, punishing “woke” Hollywood for its excessive internationalism. This intrusion of trade war into the cultural sector has pushed the geography of cinema, an admittedly niche topic, into mainstream headlines. 

As scholars working at the intersection of geography, data science, and culture, we welcome this unexpected spotlight. The sudden focus on film geography offers a rare opportunity to discuss the production and movement of films across borders, landscapes, and platforms, reflecting on the industry’s deeper cultural and economic structures. The geography of cinema is central to understanding the industry’s recent transformations and dynamics: While not all political debates about cinema are spatial, some of the most consequential, involving infrastructure, regulation, global trade, and cultural influence, are deeply rooted in place.

What’s happening to Hollywood? 

Despite some relative decline in its global dominance, the US film and TV industry remains colossal by international standards (Morris, 2024). California’s apparent vulnerability does not stem primarily from the foreign villains imagined by nationalists. The state’s share of top-grossing film productions dropped as more studios sought lower-cost alternatives. 

States like Georgia and Illinois, buoyed by aggressive tax incentives, now host a sizable share of major productions—Georgia alone accounted for nearly 20% of the top 100 U.S. films in 2022. Internationally, Canada, Australia, and the UK continue to attract big-budget projects thanks to favourable exchange rates, tax rebates, and strong production infrastructure (Sky News, 2025).

At the same time, consumer preferences have shifted too. Traditional cinema attendance in the U.S. has declined by over 25% since 2015, driven mostly by the rise of streaming services, high-quality smart TVs, and by competition from cheap content by YouTubers and TikTokers. Audiences appear less inclined to travel to a physical venue to watch high-budget films (Richter, 2024). 

[Photo by Jared Lisack on Unsplash]

The places of cinema

Place matters deeply in cinema (Anton, 2006). Films are inherently geographical: they are conceived, set, produced, and consumed in specific locales. Iconic cities like Los Angeles and New York dominate not just industry infrastructure but also our cinematic imagination (Ballatore et al, 2022). In turn, where a film is shot, set, and seen shapes its narrative and cultural influence.

Films depend on globalised value chains. On the production side, we can observe where companies operate, where talent is drawn from, where scenes are filmed, and where services like post-production and visual effects occur. On the consumption side, we can study where films are distributed and exhibited, whether in cinemas or on personal devices. Moreover, imagined geographies—whether real, fictional, or CGI-enhanced—are central to storytelling. 

For decades, film production has become increasingly globalised and transnational. Contemporary US films exemplify this global interconnectivity. Sci-fi blockbuster Dune (2021) was shot in Hungary, Jordan, Norway, and the UAE; its VFX came from a British-Indian firm, and its soundtrack was recorded in Los Angeles. Barbie (2023), a quintessentially American icon, was largely filmed in Hertfordshire, England. Conversely, films like Alfonso Cuarón’s Roma (2018) are deliberately located in a single place (Mexico, in this case), reflecting a deliberate embrace of national specificity. 

Global market, local power

The US film industry, despite the recent flight from California, is remarkably centralised. In 2024, the so-called “Big Five”—Disney, Universal, Warner Bros., Paramount, and Sony—still controlled 68% of the North American box office and are all headquartered in Los Angeles County (The Numbers, 2024). Streamers have introduced new dynamics: available in nearly every country in the world, Netflix has invested billions in producing local content across 30+ countries, hinting at a more polycentric future for film and TV production. However, while its infrastructure spans continents, crucial decision-making remains concentrated in the US and Western Europe (Lobato, 2019). Decentralisation of production has not yet translated into decentralisation of control structures.

Renationalising this sprawling industry may appeal ideologically, but it is neither practical nor strategically wise. As the US film industry relies heavily on international markets to turn a profit, not to mention its prominent role in projecting soft power, tariffs and protectionism may backfire by inviting retaliatory policies or hitting the already declining global appeal of Hollywood.

Computing film geographies

Cinema has a long-standing rapport with data science. Films are fascinating cultural objects that can be analysed based on their production context, directors, producers, actors, screenwriters, plots, scripts, dialogues, visual and sound patterns. In addition, paratexts like interviews, reviews, social media discussions, and ratings are rich data sources. Data about collaboration networks between film actors helped the rise of complexity science in the 1990s (Mitchell, 2009); recommender systems also matured through analysing film preferences in the 2000s and 2010s; box office prediction has become a classic machine learning problem. Our current research sits in this tradition: among other things, we are exploring geo-located themes in film scripts using large language models like GPT. This extends our earlier work on linguistic analysis of user narratives (Quattrone et al, 2018a), place references in plots (Ballatore et al, 2022), and spatial-cultural modelling (Quattrone et al, 2018b). 

Imagining cinema, imagining place

Whether or not Trump’s proposed tariffs are implemented, the broader forces of technological change and market-driven globalisation, now countered by US nationalism, will keep influencing where films are made and how places are represented on screen. In an era defined by data-driven decision making and high mobility of productions and audiences, computational geography offers a powerful lens on cinema’s ability to shape our spatial imagination.

References