BBC News and the July 2026 documentary Jared Leto: Hollywood's Dark Secret reported criminal sexual conduct allegations from multiple women against the actor; Leto denies the allegations. The public question that follows a belated exposé is almost always the same one: why not sooner? Artometrics asks a colder, answerable version of it instead: what was he still worth?
This report does not adjudicate the allegations, and it draws no conclusion about their truth. It tests a narrower, economic hypothesis against public box-office data and the record of four earlier cases in which an institution eventually stopped protecting a person it had shielded: Harvey Weinstein, Bryan Singer, Jimmy Savile, and Russell Brand. In each case, allegations against a living or deceased public figure are described using allegation-and-denial language, except where a criminal conviction or an official public inquiry has already settled the record.
The working idea is simple and stated plainly as a hypothesis, not a finding: institutions protect people in proportion to what those people are still worth to them, and a commercial studio pricing a bankable asset behaves differently from a broadcaster selling institutional trust. Six real box-office figures and four institutional timelines are the evidence base. Nothing here proves the idea. It sharpens the question.
RESEARCH QUESTION
Does an accused person's declining forward commercial value, and the type of institution protecting them, help explain when — not whether — that institution stops shielding them? This report compares Jared Leto's 2020s film returns against three earlier cases (Weinstein, Singer, and the broadcaster pair of Savile and Brand) to test whether "how much is this person still worth" is a legible variable in the timing of institutional silence and its end.
The charts below are built entirely from named, dated, public figures: worldwide gross against reported budget for six Leto-era films, and the dated sequence of exposure, firing, conviction, or inquiry in each of the other three cases. Where a section proposes a framework rather than a measured result, that is stated in the prose and in the chart caption.
FAST FACTS
THE LETO SLATE WAS ALREADY SOFTENING
Leto's six most recent film credits give a mixed but increasingly weak return profile. House of Gucci (2021, ensemble) returned an estimated 2.04x its reported $75 million budget on a $153 million worldwide gross, and Morbius (2022, lead) returned 2.02x on a $167.5 million gross against an $83 million budget, though Morbius is also widely remembered in trade coverage as a franchise-launch failure relative to its expectations. The Little Things (2021, lead) landed close to break-even at 1.03x, in a pandemic-affected release window that limits comparability.
The three most recent titles are all below break-even on the same trade-reported gross-to-budget measure: Haunted Mansion (2023, ensemble) at 0.79x, Tron: Ares (2025, lead) at 0.65x, and Masters of the Universe (2026, supporting) at 0.66x. Leto's role weight has also shifted from lead toward ensemble and supporting billing across this period — itself a market signal, independent of any allegation, about how studios are pricing him as a driver of a greenlight.
None of this is proof of anything about the allegations reported by the BBC. It is a measurable fact about Leto's standing as a commercial asset at the moment the documentary aired: a slate that had already stopped clearing its budgets twice over, and had already started shifting him out of lead billing.
WEINSTEIN WAS ALREADY CHEAP
Harvey Weinstein offers the clearest earlier data point for the "declining studio" half of the hypothesis. Contemporaneous reporting — MarketWatch, October 12, 2017 — documented that The Weinstein Company was struggling financially before Harvey Weinstein's sexual misconduct allegations became public that same month in The New York Times. The studio's box-office slate and balance sheet were already weak; the exposé did not create a healthy company's collapse, it arrived inside one already underway.
Weinstein was later criminally convicted in New York in 2020, with an additional conviction in California in 2022 — both now settled public record, not allegations. The sequence — a studio already cheap to lose, followed by exposure, followed eventually by conviction — is the calibration case for the rest of this report: it is the clearest instance of an institution's forward-value exposure being low at the moment its silence broke.
This is a calibration case, not a claim that financial weakness caused the exposé to happen when it did. Journalists, sources, and legal timelines move on their own logic. What the record supports is narrower: by the time the story broke, the institution protecting Weinstein had little forward commercial value left to lose.
A HIT IS NOT REHAB
Bryan Singer complicates a simple "flops end the silence" story. Singer was fired from directing Bohemian Rhapsody in December 2017, but he retained his Directors Guild of America directing credit on the finished film — and the film went on to major commercial and awards success through 2018 and 2019, a genuine hit released under his name. If forward commercial value were the whole story, that success should have looked like partial rehabilitation.
It did not function that way. The Atlantic's March 2019 report, "Bryan Singer's Accusers Speak Out," detailed multiple sexual-abuse allegations against Singer, which he has denied; after that report, his mainstream studio attachments largely stopped. A 2025 Variety report on a small, low-profile project with Jon Voight is described in that coverage as a fringe return, not a mainstream studio comeback. A retained credit and a hit film were not the same thing as continued institutional protection once a second, more detailed exposé arrived.
The lesson for the Leto comparison is a caution against reading any single commercial data point as decisive. A firing that preserves credit, or a hit that keeps someone's name on a poster, is not the same measurement as an institution's willingness to keep hiring that person going forward.
WHEN THE FIRM SELLS TRUST INSTEAD OF A PRODUCT
Jimmy Savile and Russell Brand test the second half of the hypothesis: what happens when the protecting institution is not a studio pricing a bankable asset, but a broadcaster whose product is public trust. Savile, a BBC entertainer, died in 2011 still institutionally protected. Only after his death did hundreds of abuse allegations emerge, and subsequent public inquiries — including the BBC-commissioned 2016 Dame Janet Smith Review — confirmed decades of abuse he was able to commit while shielded by the corporation. This is now settled public record via official inquiry, not an unresolved allegation.
Russell Brand's case arrived on a different institutional footing. In September 2023, a joint Sunday Times/Channel 4 Dispatches investigation reported sexual assault allegations against Brand spanning the 2000s; Brand denies the allegations. Unlike Savile inside the BBC's employ, or Weinstein inside a studio he controlled, Brand's public platform by 2023 was largely his own independent media presence rather than a single employer's payroll — a structurally different kind of "institution" to protect or abandon him.
Both cases sit outside the "declining studio asset" logic that fits Weinstein reasonably well. Savile was never dropped at all — the institution's failure was total and only corrected after death and an official inquiry. Brand's case shows how the model breaks down further once the "institution" is no longer a single employer with a P&L to protect.
TOWARD A PROTECTION THRESHOLD
This section is an editorial framework, not a statistically tested finding. The chart below places four cases on two proposed axes: whether the person's forward commercial value had already declined before public exposure, and whether the protecting institution is a commercial studio pricing a bankable asset or a broadcaster selling institutional trust. Weinstein sits at "studio, already declining" — the cleanest fit for a threshold story. Savile sits at "broadcaster, not declining" — protection that never broke on its own. Singer sits in between: a studio-adjacent case where a hit film complicated, but ultimately did not reverse, the collapse that followed a second exposé. Leto is plotted as an open question: a 2020s slate that is softening but mixed, inside a studio system, at the moment the BBC's reporting arrived.
Reading this map honestly requires holding two things at once. The pattern it shows — protection breaking earlier when there is less left to lose, and holding longer when the product is trust rather than a slate — fits the two cleanest cases well. It fits Singer only partly, and it says nothing yet about Leto beyond describing where his commercial position stood at the moment reporting on the allegations against him arrived.
LIMITATIONS
Allegations are not convictions. Except for Weinstein's 2020 and 2022 convictions and the findings of the BBC-commissioned 2016 Smith Review on Savile, every other claim in this report about Leto, Singer, and Brand is an allegation that the named person denies, and none has produced a criminal conviction as of this report's drafting. This report reaches no conclusion about the truth of any denied allegation.
There is a clear selection bias: this report can only study cases that were eventually exposed, which is not a random sample of institutional protection generally. Flops may precede, follow, or have no relationship at all to when an exposé runs; journalists, sources, and legal processes move on their own timelines that this report does not observe directly. The gross-and-budget ratios in leto-film-book.csv are trade-reported bands from outlets such as Box Office Mojo, Deadline, and SlashFilm rather than audited studio accounting, and are labeled here as derived figures, not confirmed net financials. The "protection threshold" framework itself is an editorial hypothesis map built from four cases; it is not a statistically fitted model and should not be read as one.
CONCLUSION
The honest version of this report's finding is modest. Jared Leto's 2020s film slate had already softened — three straight titles below break-even on trade-reported figures — by the time the BBC's July 2026 documentary reported allegations against him. That fact alone proves nothing about timing or motive. What the four comparison cases add is a sharper question rather than an answer: institutional protection appears to track forward commercial value more legibly when the institution is a studio pricing a bankable asset (Weinstein) than when it is a broadcaster selling trust (Savile), and a single hit film is not proof that protection has resumed (Singer). Where Leto's case falls on that map is, for now, an open question this report is built to keep asking rather than to close.
DATA AND METHOD
The film-return figures come from docs/content-os/drafts/protection-threshold/leto-film-book.csv, six rows covering Jared Leto film credits from 2021 through 2026: title, release year, role weight (lead, ensemble, or supporting), worldwide gross in millions of US dollars, reported budget in millions of US dollars, and the derived gross-to-budget ratio. Gross figures are drawn from Box Office Mojo; budget figures are trade-reported bands compiled from SlashFilm and Deadline coverage of each title, and are treated throughout as derived estimates rather than audited studio financials. The gross-to-budget ratio is a derived metric calculated as worldwide gross divided by reported budget; it does not account for marketing spend, distribution fees, or backend participation, and is not a substitute for studio profitability accounting.
The institutional comparison cases (Weinstein, Singer, Savile, Brand) are built from named, dated public reporting and official inquiry findings rather than a structured dataset: MarketWatch's October 2017 report on The Weinstein Company's finances, Harvey Weinstein's 2020 and 2022 criminal convictions, The Atlantic's March 2019 report on Bryan Singer, Variety's 2025 report on Singer, the BBC-commissioned 2016 Dame Janet Smith Review on Jimmy Savile, and the September 2023 Sunday Times/Channel 4 Dispatches investigation into Russell Brand. The "protection threshold" chart in the Toward a Protection Threshold section is an editorial index built by Artometrics from the facts above; it is not derived from a quantitative model and its axis positions are qualitative placements, stated as such.
REFERENCES
BBC News. Jared Leto criminal sexual conduct allegations, July 2026. https://www.bbc.com/news/articles/cd7lg2nz2x2o — Primary reporting on the documentary's claims and Leto's denial; allegation language only.
BBC Three. Jared Leto: Hollywood's Dark Secret. https://www.bbc.co.uk/programmes/m002zl1y — Documentary programme page.
Box Office Mojo. Film gross data. https://www.boxofficemojo.com/ — Observed worldwide gross figures for the Leto 2020s slate.
MarketWatch. The Weinstein Company struggled financially even before Harvey Weinstein's sexual misconduct allegations, October 12, 2017. https://www.marketwatch.com/story/the-weinstein-company-struggled-financially-even-before-harvey-weinsteins-sexual-misconduct-allegations-2017-10-12 — Contemporaneous reporting on TWC's pre-exposé financial decline.
The Atlantic. Bryan Singer's Accusers Speak Out, March 2019. https://www.theatlantic.com/magazine/archive/2019/03/bryan-singers-accusers-speak-out/580462/ — Report after which Singer's mainstream studio attachments largely stopped.
Variety. Bryan Singer secret movie, Jon Voight, 2025. https://variety.com/2025/film/news/bryan-singer-secret-movie-jon-voight-israel-lebanon-1236499696/ — Reporting on a fringe project, distinguished from a mainstream studio comeback.
Related Artometrics reports: The Prestige Economy: Oscars, Grammys, and Nobels.
EDITOR'S NOTE
This is a reproducible Artometrics report: the underlying film-return data ships as a downloadable CSV via the site's Download menu. Chart PNG exhibits for the figures above are pending the standard Artometrics render pipeline and will be attached alongside the live Plotly views once generated. This report makes no finding about the truth of any denied allegation described above; where a matter is settled public record via criminal conviction or official inquiry, that status is stated explicitly in the text.