Methodology note

Companion methodology note to Sulco Study No. 2, Indonesia’s cinema gap (30 Oct 2026, v1.0). Figures here are the same set used in the deck, the summary and the study page; each headline number appears once on the Canonical sheet of the study workbook with its source, grade and confidence.

Results

Indonesia has 8.4 screens per million people against about 18.5 predicted at its income, and each screen sells about 53,100 tickets a year. Four paths to 31 December 2030 give 81 to 200 million tickets a year. No probability is assigned to any of them.

The question and the decision

Is Indonesia under-screened (demand is there, supply is not), under-demanded (people do not go), or is the 2025 number concentrated in a few cities and a few films? And what would it take, and what would it be worth, to look like its neighbours?

The decision the study informs: whether Indonesia’s cinema shortfall is a supply problem worth capital and policy. Readers: exhibitors and their investors, producers, the culture and creative-economy ministries, and regional governments that set the entertainment tax. Primary horizon: 31 December 2030.

Four rival explanations and the test that separates them: (a) under-screened: high tickets per screen with normal tickets per person (Tests 1 to 3); (b) under-demanded: low tickets per person against income and price; (c) a two-hit artefact: concentration and the median film; (d) streaming substitution: subscriptions against admissions and release windows.

How the industry is counted

Indonesia publishes no official box office. Badan Perfilman Indonesia (BPI) counts screens and cinemas (2,375 and 496 in November 2025); the Ministry of Culture counts tickets to Indonesian films (80.27 million in 2025); private trackers (Bicara Box Office, Cinepoint) estimate the total including imports (126.2 million in 2024). Every ratio in the study says which count it uses. The 2025 total, about 121 million, rests on one industry estimate and carries grade C and low confidence, so it appears only in body text.

The cinema census

Sulco counted 518 cinema sites on 29 and 30 October 2026: the Cinema XXI locations list (270 sites), CGV’s cinema list (71), Cinépolis and small-chain tables with press citations, and showtime listings for independents. Sites without current showtimes were treated as closed. Each site was matched to one of Indonesia’s 514 regencies and cities, with BPS 2025 populations. A regency counts as covered if it has at least one cinema; coverage is by address, not travel time. Result: 170 regencies and cities with a cinema, covering 62% of people; 344 without, home to 109.1 million.

The six measures

M1 is the headline measure, with M4 beside it: M1 is the number a reader can picture, and M4 shows the screens that exist are full. M3 and M4 rest on tracker admissions; M5 multiplies them by one chain’s ticket price.

Tests against income

Test 1, screens. Screens per million on the log of GDP per person (PPP, constant 2021 US$, WDI 2024) across 36 countries: R² 0.58, each doubling of income goes with about 19 more screens per million. At Indonesia’s $14,500 the line predicts 18.5; Indonesia has 8.4, a gap of about 10. Thailand (13 below), Malaysia (11) and Vietnam (8) also sit below the line. Without Ukraine and Kazakhstan (weak data) the line predicts 19.4.

Test 2, tickets. Tickets per person on income across 35 countries: R² 0.52. Predicted 0.52 at Indonesia’s income; actual 0.45 (2024, tracker). On screens Indonesia is about half its line; on tickets about a seventh below. A Rp46,000 ticket is about 21% of a day’s GDP per person, like Vietnam (24%) and India (23%).

Both lines are descriptive associations, not causes.

Scenarios and assumptions

Four scenarios to 31 December 2030, built with formulas off the Assumption_Register sheet. Population grows 0.8% a year. A: screens grow 2.5% a year (the 2019 to 2025 pace) and tickets per screen stay at 53,100. B: screens reach 15 per million (between Vietnam and Thailand). C: tickets per screen fall to Malaysia’s 30,000. D: tickets per screen fall 20% as the local-film run fades. Box office at the 2025 Cinema XXI average ticket of Rp46,057.

No probabilities. The house method allows odds only where a base rate supports them. The study has three growth stories (Vietnam, Korea, Thailand), not a reference class written before counting, so the scenarios are a range, not a forecast.

Assumption register

Industry size at 200 million tickets

Spending at the cinema (tickets and food) rises from Rp9.1 trillion to Rp14.4 trillion a year at 200 million tickets, still about 0.07% of GDP. Entertainment tax to cities rises from Rp0.5 to 0.8 trillion; the producer side, after the distributor fee, from Rp2.3 to 3.7 trillion; direct cinema jobs from about 8,900 to 16,600 at 3.7 staff a screen. Building 2,065 screens at Rp5 to 8 billion each needs Rp10 to 17 trillion. Sensitivity: ticket price 10% either way gives Rp13.5 to 15.3 trillion of spending; an exhibitor share of 45 to 55% gives producers Rp3.3 to 4.0 trillion.

Confidence and the case against

Overall confidence: moderate. The strongest case against our reading: totals are tracker estimates, and MPA puts 2024 box office at US$294 million, 25% below Cinepoint’s US$392 million. At MPA’s level each screen sells about 40,000 tickets, near Vietnam’s 49,000. New screens outside big malls may not pay: CGV is loss-making and parent-funded, Cinépolis shrank, and the census found closed small-format sites. The boom is two films deep: two titles took 26% of 2025 local tickets.

Our reading after weighing it: the shortage holds even at MPA’s lower count, since 40,000 tickets per screen is still a third above Malaysia and Thailand. The build-out case is weaker than the arithmetic: regency screens must be cheap to build and run. What would settle it: an audited national admissions count, ticket sales per screen at regency sites, and 2026 totals on a normal film slate.

Data gaps

• No official box office or total admissions: every per-person ratio rests on trackers.
• The early-1990s peak: 2,853 screens (archival series) against 3,300 and 6,600 in other accounts.
• Malaysia admissions: 48.7 million (Comscore 2023) against 32.6 million (FINAS 2024).
• Thailand has no official national screen series; the Philippines has published none since 2020.
• Producer credits: co-productions are credited to every partner.
• The exhibitor split and distributor fee are conventions, not reported terms.
• Cinema XXI screens per site: the coverage map uses a chain average of 5.17 for 270 sites.

Canonical numbers

The Canonical sheet lists every repeated number once. Grades: A primary and audited, B official or reported by a named body, C a single estimate. Sources: 21 graded A, 27 B and 11 C.

Review status

The study follows the Sulco Research Methodology (house method) v1.0, which took effect on 30 October 2026, the day the study started. Two steps could not be done in order and are recorded here rather than hidden: the framework was written on day one but not frozen as a timestamped record before data work, and the pre-mortem review gate was not run before the data. An independent number audit traced 54 figures and all required fixes were applied. The numbers were checked by the author's own review before publication. Review by an outside expert (an exhibitor, producer or film-policy academic) had not taken place at publication; it is planned before the next update on 31 January 2027.

Change log