What A Century (Plus a Pandemic) Does to Moviegoing and Why It Matters
Over the last six months, there has been a lot of talk about the future of theatrical exhibition (“going to the movies”), with The Discourse treating every film as some lesson on what does or doesn’t work, what’s dead or “so back,” Hollywood’s hubris or the audience’s oversight, and so on. There is no practical way to assess the current state of the North American box office without considering the last century of moviegoing – during which there are perhaps four different acts. The third, which began shortly after the millennia and ran until the COVID-19 pandemic, is the best place to start.
Act Three (2000-2019)
The domestic box office has been in persistent decline since 2002, when per capita annual admissions in North America hit 5.1, their highest point since 1977’s Star Wars. By 2019, the last pre-pandemic moviegoing year, per capita admissions had plummeted to less than 3.5 – not because fewer people were going to the movies (74% of Americans bought a ticket in 2002, and 76% did in 2019) but because the share who went at least once a month on average fell by nearly two thirds (from 28% to 11%).
There are various explanations for this decline. One argument, which has been passionately argued by Quentin Tarantino, is that rising ticket prices have priced out the working class. Superficially, this theory might seem compatible with the frequency data — the average person can still afford to go, but only 3-4x per year. Yet while ticket prices grew quickly in the 2000s, they added only a dollar in inflation-adjusted terms over twenty years and were no costlier than in the 1980s and actually cheaper than in the 1960s and 1970s (tickets today also buy much better screens and seats, plus reserved seating, none of which directly rebuts the affordability argument, but also means value per dollar has increased). And if theaters were deliberately pricing out lower-income consumers in order to increase pricing among the wealthy, then one would expect significant net revenue growth. Instead, gross revenues grew by 23% over 16 years, or a meager 1.28% annually. On an inflation-adjusted basis, revenues fell 13%, or −0.95% per year. In contrast, the US economy grew at an average annual rate of 4% in real terms over that same period. It’s also worth highlighting that since 2002, all other forms of spectator amusements (theater, music, sports, etc.) have seen even greater ticket price increases, yet attendance soared, and inflation-adjusted revenues grew 2.9–5.1% annually.
To explain the persistent underperformance of the box office against other entertainment benchmarks, the blame gets spread around: Hollywood’s obsession with sequels, franchises, blockbusters, and IP (“SFBIP”); related disinterest in the adult dramas, R-rated action movies, PG 13 rom-coms, all sorts of comedies, and other sorts of original bets that routinely topped the charts in the 1970s through the 1990s; and the parallel rise of SVOD services. The last point is easiest to address as the decline in attendance was steepest before Netflix, Hulu, and Prime Video debuted (and it took several years for these services to premiere their own original films online). Moreover, the largest share of the decline (which was the sharpest since JFK was elected president) predates Hollywood’s overall shift to SFBIP, too. It wasn’t until 2010 that Pixar released its second-ever sequel (which was also 11 years after its first), at which point admissions had already fallen by more than a ticket. The Avengers premiered two years later (2012) and at that point, only five films had released over five years (more recently, the MCU releases 3-4 annually). The first Disney Star Wars film premiered two weeks before the start of 2016. The DC Extended Universe didn’t really begin until 2016, either.
In 2019, 900 films were released in theaters, up from 350 at the start of the millennium. Essentially all of this growth came outside the six majors (including their indie labels), as the majors continued to release about 100 films collectively per year. Yes, these “tail” films released to fewer screens than those of the majors (though exhibitors would happily have shown these titles if the audience demand warranted it), but it’s notable that the majors’ share of total ticket sales grew from the low 70s to the mid-80s despite their declining share of total releases and the arguments that their focus on SFBIP was turning away customers.
Indeed, the entirety of the decline in per capita consumption from 2002 to 2019 came from films outside the top 25, and those top 25, and as we know, those films were increasingly, well, SFBIP.
The clearest disproof of the SFBIP-killed-the-box office theory comes via demographic research (unfortunately, the data is not available earlier than 2009). The average American over two years old bought 4.31 tickets in 2009 and 3.46 in 2019, or 0.85 (−20%) fewer tickets. However, those aged 2–11 dropped by 1.2 tickets (−29%), 12–17 fell 3.00 (−38%), and 18–24 was 3.7 (−44%!). Certainly, the paucity of raunchy comedies has reduced movie attendance among teens and twentysomethings, but I’m personally skeptical that 2–24-year-olds would come out in greater force if only there were more adult dramas, original tales, and “cinema.” Moviegoers in those ages group used to go the theaters so often because it was an essential venue for babysitting, socializing, and dating (and later, going shopping in the attached mall). The “job” of the movies has been substantially displaced in each of these areas over the last twenty years.
And overall, older cohorts were, before COVID, attending the movies as much as ever. In fact, 40–49s were up! I’d even argue this increase is understated and that, depending on how you look at it, 25–39s probably increased too. Why? Well, 2-to-14-year-old customers (roughly 15% of tickets) don’t attend the movies by themselves; they need a parent, grandparent, or guardian who is likely 25-plus and in many cases is attending the film for the child rather than for themselves. As such, every ticket a 2–14-year-old doesn’t buy represents a lost purchase for a patron in one of the four oldest customer tranches; therefore, the stability of 25–39, 40–49, 50–59, and 60-plus purchasing rates means that those customers chose to go to movies for themselves more in 2019 than in 2009 (though again, these new tickets went overwhelmingly to SFBIP films). Even without this adjustment, the observed durability in purchasing across these segments refutes the common arguments for the box office’s post-2000 decline: the proliferation of large high-definition TVs and high-quality audio systems, the availability of on-demand films through TVOD and SVOD, and — yes, yet again — the quality or attractiveness of the films produced.
Act Four (2020-)
After COVID, concerns about the viability of the box office intensified. With the advent of vaccines and related trends, the world had begun to open up, true, but it also seemed that the pandemic had forever impaired the theatrical business or, at a minimum, accelerated its otherwise slow secular decline. Consumers “benefited” from a few years where signature feature releases (Dune, Wonder Woman, Marvel flicks, Pixar originals) debuted day-and-date on streaming services, typically at no additional cost. Those films that did have a theatrical window also had a short one — as little as three weeks. Meanwhile, many consumers still couldn’t go to the theater. The collective result was the emergence of new behaviors (new movies at home!) and a temporary cessation of the old behavior (going to the movies!). Domestic per capita admissions, which had shocked when it fell under the 3.50 mark in 2019, fell to 0.62 in 2020 and 1.26 in 2021.
The fears of permanent impairment were partly allayed by 2022 and 2023, which benefited from relatively dense release schedules, as the many films that were supposed to be released during the pandemic finally debuted on the Silver Screen. Top Gun: Maverick, which was delayed for two years at an estimated interest expense of a million a month, became the biggest domestic release of 2022 (it grossed over $1MM a day for 75 days) and the second-largest worldwide. 2023, meanwhile, had Barbenheimer. The worldwide box office of 2022 beat 2021 by 27%, and 2023 beat 2022 by 31%. Domestic per capita admissions were still but 2.31 — down a third from 2019, sure, but seemingly on the mend.
But 2024 has started to pay that back with a record (non-pandemic) decline, with the domestic box office after 24 weeks down over 25%. There was some wishful thinking that the decline of the Marvel Cinematic Universe — which at its peak was producing 3–4 films per year, each of which would devour 5% of America’s scant and effectively capped ticket volumes — might allow some other films to thrive. But it’s more likely that Marvel’s decline has cost the market well more than it has shared with rival films.
Altogether, 2024 is currently trending for roughly 1.8 tickets sold per American, fewer than in 2022. Even if 2025 rebounds to 2.5 or so (which would beat 2022 by 8%), it looks like the pandemic will have accelerated the decline in per capita theatrical admissions by roughly 15 years. Pre-pandemic, the industry was trending toward 2.75 by 2029.
Due to the success of films such as Oppenheimer, Barbie, and Sound of Freedom as well as the struggles of Marvel throughout 2023, some have argued that audiences were finally seeking out alternatives to the tired plotlines, superheroes, and franchises that dominated the last decade. According to these theories, post-pandemic attendance is primarily impaired by post-pandemic film slates, and once Hollywood adjusted to new audience preferences, attendance would soar again. Perhaps not to the levels of the early 2000s or even 2010s, but at least back to 3 tickets per person per year.
The early evidence suggests otherwise. To start with, 2023 was not the rejection many made it out to be: The #1 film, Barbie, is an adaptation of a 64-year-old toy that has generated more franchise revenue than all but a half dozen properties ever and is instantly recognizable around the world and beloved to many. #2 is based on arguably the most iconic video game character (all I need to say it’s “It’s-a-me”), who is also in his fifth decade. #3 was the animated Spider-Man: Across the Spider-Verse. For decades, Spider-Man has been the bestselling and most popular character based on a comic book, and the character has over 10 appearances in the hundred highest grossing films at the domestic box office. #4 was the 32nd feature film in the Marvel Cinematic Universe, Guardians of the Galaxy Vol. 3, which beat 2017’s Guardians of the Galaxy Vol. 2, which ranked 5th, and was only one slot lower than 2014’s Guardians of the Galaxy (i.e., Vol. 1). #5 was Oppenheimer, a remarkable outlier, yes, but #6 was Disney’s live-action remake of their own 1989 film The Little Mermaid (which ranked 13th at the time). #7 was (Willy) Wonka, Warner Bros.’ de facto prequel to the 1971 film, and which Disney had remade into the 7th biggest film of 2005. #8 was the 31st MCU film, Ant-Man 3, which beat its 2018 predecessor (9th place), and the original Ant-Man 1, which was 14th place in 2015. #9 was John Wick 4 (John Wick 3 was 14th in 2019, while John Wick 2 was 35th in 2017, and the original was 77th in 2014), with Sound of Freedom coming in #10. And the second Avatar film, released only two weeks before the start of 2023, is the third-highest-grossing film ever and grossed nearly half its total during calendar 2023. Personally, I wouldn’t call this a breath of fresh air (which doesn’t mean it was stale).
Moreover, the main reason 2024 seems weak is the relative paucity of SFBIP releases. To this point, it’s the first year since 2012 that the MCU will release only a single title – and this title, Deadpool 3, is expected to be the year’s highest grosser even though (or possibly because!) It’s the 34th entry in the MCU (and the 73rd Marvel-related movie total). It will probably earn at least a third more than 2016’s Deadpool or 2018’s Deadpool 2. Some of 2024’s shortfalls comes from strike-related delays, which bumped films like Pixar’s Elio, Disney’s Snow White, and Marvel’s Thunderbolts into calendar 2025, though labor action also moved films like Dune (Part 2) into 2024, and 2024’s relatively thin release calendar should have aided titles such as Fall Guy and Furiosa. We’ll never know what 2023 and 2024 attendance would have looked like without the strikes, but per capita admissions are so far off the pre-COVID trendline that the strikes are, themselves, far from a sufficient explanation. And in general, the relevance of a year’s slate also tends to be overestimated; 2019 was perhaps the greatest blockbuster year in history, yet tickets sold fell 6.5% year-over-year.
Annual slates will probably produce more year-to-year variability in a post-pandemic environment, but to that end and the relevance of SFBIP to moviegoing, the primary hope for 2025 is, indeed, its very strength in SFBIP titles. This slate includes the Minecraft movie, Superman, The Batman Part II, Mission: Impossible 8, Fast X: Part 2, the live-action adaptation of How to Train Your Dragon, Jurassic World 4, Marvel’s Captain America 4, the live-action adaptation of Snow White (the top movie of 1937), Marvel’s Thunderbolts, the live-action adaptation of Moana (#13 in 2016), Marvel’s The Fantastic Four, Marvel’s Blade, Zootopia 2 . . . and, of course, Avatar 3, which will probably be one of the five biggest movies ever.
No slate since 2019 seems more likely to return those sub-24-year-olds to the box office. And yet these customers (47% of tickets sold in 2009 before accounting for parents) will probably never return in full force. There are too many other (and typically more social) media alternatives, ranging from TikTok to Fortnite. Those under ten seem particularly compromised. Inside Out 2 is an extraordinary feat, but it’s inconceivable that annual attendance among 2–11 year olds will ever match the levels experienced before the Q4 2019 launch of Disney+, which includes, among other titles, dozens of Pixar and Walt Disney Animation films, each one infinitely playable at no added cost, and at home.
Preorder “THE METAVERSE AND BUILDING THE SPATIAL INTERNET,” the fully revised and updated edition of my nationally bestselling (US, UK, Canada, China) and award-winning (Best Books of 2022 year by Amazon, The Guardian, FT China, The Economist’s Global Business Review, ByteDance, Barnes & Noble) title. A limited run of signed copies can be secured here.
Acts One (1900-1960) and Two (1960-1999)
Before getting to what might be done, it’s helpful to look farther back than just the post-Y2K history of the box office. Why? Because in 2024, more movies will be made than ever, more movies will be watched, and total revenues will be greater than ever. Yet if theatrical attendance figures were shared with a Hollywood executive during the Golden Age, it would be impossible to imagine the industry could even exist.
For close to 30 years, the average American went to the cinema 20–30 times per year. Beginning in the late 1940s, attendance began its secular decline, largely due to the rise of broadcast television (per capita admissions fell below 20 just after The Ed Sullivan Show debuted), and cable TV after that, before stabilizing at 4–5 tickets per capita per annum in the 1970s (around the time color TV reached 90% penetration in the United States). Contrary to popular belief, the rise of VHS in the 1980s and 1990s had little effect on theatrical attendance overall, even as it became an equally lucrative distribution channel.
At its peak in the late 1940s, moviegoing represented an astonishing 1.3% of all personal consumption expenditures in the United States, or $1.3 of every $100 spent (PCE refers to the “Total amount of money spent by households and non-profit institutions serving households on goods and services, during a specified period. This includes expenditures on durable goods, such as cars and appliances, non-durable goods, such as food and clothing, and services, such as healthcare, education, and recreational activities”). From 1930 to 1947, theatrical’s share of total PCE was consistently above 1%. By 2019, this share had fallen from $1 in every $100 to 9¢ (a 95% total decline that contrasts with the roughly 82% drop in attendance).
It was inevitable the theater’s share of PCE would decline as U.S. economy modernized and became increasingly wealthy throughout the 20th century; there were too many new things to do. Indeed, some of this spend went to TV which was then used to watch movies. By the mid-2010s, HBO (née Home Box Office), the most profitable network (by margin) in the world, reported 40% of its subscribers only watched movies on the service, rather than its outstanding and famous original TV series, and that movies comprised over 60% of total viewing, too.
These points provide important context, but they’re also a bit misleading. Even among ticketed spectator amusements – live entertainment (which is mostly concerts but also includes performance theater, ice shows, etc.) and live sports – moviegoing has lost share. In the 1920s to 1950s, motion pictures were $80 or more over every $100 spent on a spectator amusement in the Unite States. Today, it’s less than $15.
Not only have these other ticketed spectator amusement categories taken share from moviegoing, but they’ve also grown considerably overall. And to provide the fairest comparison, we can narrow in on growth since 1960 – at which point per capita admissions had fallen to eight annually.
In the 60 years from 1959 to 2019, the theatrical movie business has grew from $9.8 billion (in 2023 dollars) to $13.4 billion, a gain of 33% or $3.3 billion. In contrast, the ticket live entertainment (excluding sports) business grew from $2.9 billion (in 2023 dollars) to $47 billion, a gain of 1,500%, or $44 billion (in other words, it went from 29% of the size of theatrical moviegoing to 350% the size). Live sports grew from $3.3 billion (in 2023 dollars) to $37 billion, a gain of 1,050%, or $34 billion (in other words, it went from 32% of the size of theatrical moviegoing to 280% the size). And whereas theatrical moviegoing is down a third (in 2023 dollars) from 2019 to 2023, live entertainment is up 2% and live sports is down 8%.
The growth in these other spectator categories also helps breakdown some of the longer running explanations for theatrical moviegoing’s decline. For example, improvements in at-home equipment (e.g. bigger and better screens, audio systems) apply to live sports just as they do to feature films. Despite this, the number of sporting events has grown, average attendance has grown, and average spend has grown. In 1960, the average American spent $21 (in 2023 dollars) attending live sporting events each year. In 2019, that figure was $117. And unlike going to the movies, physically attending a sporting event often provides a demonstrably worse viewing experience. In exchange, audiences receive an elevating shared spectator experience – but this is supposed to be the case for theatrical moviegoing too (albeit less powerfully).
Unlike live sports or theatrical moviegoing, live music is insulated from improvements in at-home audiovisual equipment. Most concerts aren’t broadcast and when they are, the experience is fundamentally dissimilar. At the same time, industry has done the best job of elevating and enhancing production, while driving up pricing (like sports, partly through increased customer discrimination). Per capita spend was $17 (in 2023 dollars) in 1960 but $148 in 2019. Inflation-adjusted movie ticket prices in 2024 are unchanged from the 1960s and 1980s even though screens are much better, seats far more comfortable, film budgets have skyrocketed, and attendance has gone from every two-to-three months to twice or so a year.
The Foreign Interlude
North America’s declining interest in going to the movies is not unique – not over the last 25 years, or the last century. It just seems that way due to the market’s historical appetite for the cinema, role pioneering the artform, and economic maturity.
From 2009 to 2019, the international box office grew by $11.5B, but 75% of this growth came from China. The country’s surging revenues came not from an increased demand for films, per se, but the establishment of the moviegoing business in the first place. In 2000, China had only a few hundred movie screens (and only 3% of its population classified as “middle class,” whereas today it boasts over 80,000 screens (twice that of the United States) and more than half of the population is middle class. In more mature markets, growth has been modest-to-flat (or in inflation-adjusted decline) for a decade or two. Box office revenues in EMEA are flat over the last fifteen years, for example, and the Latin American box office has been flat since 2011. From 2009 to 2019, APAC (ex-China) grew only 25% in nominal terms.
It’s also worth noting that in most foreign markets, demand for theatrical moviegoing was never as high as it was in North America. In 2002, the average American went to the box office 5.1 times, but Germans attended roughly 1.5 times, Swedes went 1.9 times, U.K. and French residents 2.5 times, and so on. Interestingly, these markets have seen modest (and sometimes no) decline up until the pandemic, which suggests some uniform “Western” baseline for attendance that Americans long exceeded. Unsurprisingly, the box office in international regions were down between 1-15% in 2023 versus 2019, while the US box office was down 21%.
And as I detailed in my 2022 essay What China, Marvel, and Avatar Tell Us About the Future of Blockbuster Franchises, the modernization of foreign markets tends to transform their habits. China is an exaggerated, but still instructive case (Nigeria, India, and France are also fascinating). Through a mixture of censorship and national investment strategies, the country rapidly weaned itself off of foreign films and built-up domestic alternatives. From 2011 through 2013, for example, an average of 5.3 of the top 10 films at the Chinese box office were Hollywood-made. From 2021 through 2023, only 1.3 were, with none hitting the top 10 in 2023. That revenue stream isn’t gone, but it’s certainly tiny compared to the 2010s (it also nets Hollywood studios barely a quarter of the ticket price paid, rather than the roughly 40% internationally and 50–60% they get in the United States).
Screentime
Moviegoing is old. The first halls appeared over 125 years ago, with modern-styled theatrical palaces emerging in the 1920s, talkies rising to prominence in the following decades, and color not long after, after which per capita annual admissions rebounded from the low 20s to low 30s. In the seventy years since, there have been many technological advances – Cinerama, Colorama, Panoramic Screens, Cinemascope, 3D, Stereophonic Sound, IMAX, ScreenX, etc. – which, for some or all moviegoers, improved the moviegoing experience. However, they have not fundamentally renewed interest in the medium (even though many of these same technologies made concerts, performance theater, and live sports more appealing).
In 2024, the domestic box office will be in its 22nd year of sustained decline. And due to the pandemic, audiences are behaving as though they’re between 32 to 37 years into this decline. Fewer than two thirds of Americans still go to the movies, and on average, they will purchase just about 3 tickets annually (hence the average American buying about 2). The practically addressable number of tickets is even more modest as a handful of signature releases each year (e.g. an Avengers, Jurassic, Avatar, Despicable Me) will devour 5-10% each. These constraints mean that the box office – audiences – won’t support many films, or many great films. The misses will consistently surprise moviegoers, critics, stars, and reviewers. This is not a new challenge, per se, but it has never before been more brutal (note that while the modern dominance of comicbook movies is often likened to the heydays of Westerns, Westerns thrives at a time where Americans headed to the theater 20-35x a year!). This will have to change budgets, talent incentives, risk proclivities, franchise plans, and more. But while annual attendance is probably more sensitive to overall slates than in decades (pre-pandemic, there was perhaps ±2% variability in total ticket sales by year), it’s pretty capped.
Still, changes are probably due. An independently operated MoviePass was always a dumb idea, but to renew frequent moviegoers, it’s clear that some form of AYCE subscription or subscription perk will be required. AMC A-List is a good start, but doesn’t Disney+ have an additional tier (perhaps Disney++) that provides free or discounted tickets to Disney films while they are in theaters? For that matter, distributors should sell premium movie tickets that include EST entitlements or discounts (this may not increase attendance, but it should increase total revenue per customers).
Some executives say ticket prices must go way down. It’s more likely they should substantially increase to reflect the scarcity of moviegoing for the average moviegoer (it stands to reason that if the average American has dropped three of five trips per year that the remaining two should be more, not less valuable on average. Of course, exhibitors don’t want these increases to cannibalize moviegoer’s spend at the concession stand, none of which is shared with distributions and all of which is high margin. Perhaps a new deal needs to be worked out which better aligns incentives and the concession behavior of artists.
Although a few observers fear that seat reservations have deprived the industry of much-needed in-theater trailer views. I suspect trailers need to be cut (way) back. Watching 25 (sometimes 30!) minutes of trailers before the 83 pre-credit minutes of the Super Mario Bros. Movie is crazy (especially if you have restless kids). Watching 25 minutes before the 170 pre-credit minutes of Oppenheimer is painful. Those with smaller bladders inevitably end up stepping out of the film they paid for in part because they had to watch two dozen minutes of trailers first.
The industry is understandably reluctant to give up any opportunity to pitch a captive moviegoing audience on new movies, but with so few tickets bought by moviegoers (one of which is literally underway), it’s likely just an obstacle to going to the movies in the first place and cumulatively deprives exhibitors of extra showtimes per day (which are increasingly scarce due to movies getting longer). High trailer loads might have made sense in an era where trailers didn’t dominate social media discourse and YouTube trending pages (or amass hundreds of millions of views in the first 24 hours online), but now it’s a bother. Three trailers should suffice.
Exhibitors and distributors would probably benefit from allowing more “relaxed decorum” showings, too. Taylor Swift: The Eras Tour, which ranked 11th in 2023, soared in part because fans could get up and dance (including on the frontal stage) and even record the film. Similar showings should be available for The Avengers, Despicable Me, and more.
Some recent changes should probably be unwound, too. While rapid PVOD windows have helped some money-losing films recover their investments, this model probably just trains audiences to skip uncertain releases because they might be available at home in three weeks anyway.
With or without these changes, the century-plus history of moviegoing – and the partial, if as-yet still disappointing post-pandemic rebound – should give industry observers and participants some comfort. North American moviegoing is on the decline, yes, but it will endure for the foreseeable future. Maybe that’s 2, 2.5 or even 3+ tickets per person per year, but that’s enough for $10B (and maybe much more as the population and prices grow).
Mostly, though, the industry needs to continue to look beyond the Silver Screen – and much farther than SVOD licensing, too. That’s the goal of the second entry in this two part series, which premieres next month.
Matthew Ball (@ballmatthew)
