G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G: Yes. I think I can turn what you're asking into something measurable.

Your counterfactual is:

Assume Donald has no personally damaging Epstein material whose concealment would rationally justify the administration's behavior. What has the United States paid—in scarce congressional attention—because the administration did not simply conduct a competent, prompt, legally compliant release?

That's importantly different from asking how much time Congress has spent investigating Epstein. Some of that investigation could be worthwhile regardless. You're asking for the incremental congressional cost created by the release/obstruction controversy itself.

I did a first-pass reconstruction, and the answer is surprisingly large.

The cleanest number: probably thousands of member-hours

I would measure this in member-hours, analogous to person-hours: one member of Congress occupied for one hour = one congressional member-hour. That's much more meaningful than saying "Congress spent 40 hours," because a two-hour committee proceeding involving 30 members consumes vastly more legislative capacity than one senator giving a two-hour speech.

There are several layers.

First, there's the original legislative fight over disclosure. The House Rules Committee had to hold proceedings over an Epstein-release resolution in July 2025. Then House leadership effectively shut down legislative business and began the summer recess a day early rather than allow the Epstein dispute to continue generating votes. citeturn1search4turn1news69 That last one is potentially enormous: you shouldn't count 435 × 8 hours because members weren't necessarily going to spend every minute legislating, but even assigning a very conservative fraction of that lost day to displaced House business produces hundreds of member-hours of opportunity cost.

Then Congress eventually had to pass an actual statute telling DOJ to release material that, under your counterfactual, the administration had no important presidential reason to withhold. H.R. 4405 received 40 minutes of House floor debate and a 427–1 vote; the Senate then passed it by unanimous consent without debate, and it became Public Law 119-38. citeturn1search10turn1search6turn1search7 Just the House floor debate, if roughly 428 voting members were being held for that business, represents on the order of 285 member-hours, although that's an intentionally crude conversion because not every member was physically listening to all 40 minutes.

But passing the law didn't end it.

DOJ's implementation produced another congressional oversight cycle. By March 2026 the House Oversight Committee had actually voted to subpoena Attorney General Pam Bondi over DOJ's handling of the files and compliance with the Transparency Act. citeturn0search0turn2news55 And by July, legislators had introduced an Epstein Files Transparency Act II, specifically designed to create judicial enforcement mechanisms for alleged DOJ noncompliance with the first law. There are parallel House and Senate bills. citeturn0search3turn0search4turn0search7

That's a remarkable fact for your counterfactual. Congress spent legislative capacity passing:

"Release the files."

and subsequently spent more legislative capacity considering:

"No, seriously, we meant it, and here's a mechanism for enforcing the first instruction."

Meanwhile, the House Oversight investigation ballooned. It subpoenaed DOJ, the Epstein estate, Maxwell, Bill and Hillary Clinton, former attorneys general and FBI directors; conducted interviews/depositions with Barr and Acosta; held a private survivor roundtable; pursued financial records; and continued interviewing witnesses well into 2026. citeturn2search3turn3search7 By this summer the committee was releasing transcripts from Noel, Bondi, Groff, Gates, Black, Band, Ruemmler and Staley, among others. citeturn3search1turn3search2turn3search3turn3search9turn3search5

Some of that cannot fairly be charged to Donald's handling. Congress investigating Epstein's crimes, institutional failures, trafficking, financial networks, or how prosecutors treated Epstein is legitimate independent oversight. We need to subtract it.

But there is a very useful dividing line:

Count activity whose necessity disappears if DOJ had promptly made a competent, lawful release.

That includes fighting over release resolutions; Rules Committee proceedings over forcing disclosure; discharge/release maneuvering; floor debate on legislation compelling disclosure; congressional demands concerning DOJ compliance; Bondi compliance proceedings; briefings about redactions and withheld documents; legislative work on EFTA II; and some portion of the enormous staff effort spent repeatedly determining whether DOJ had actually complied.

It does not automatically include interviewing Acosta about the original plea agreement, interviewing people with substantive knowledge of Epstein's network, investigating trafficking failures, or talking with survivors.

And here's where I think your intuition lands.

My preliminary range

I'd currently be comfortable saying:

Direct, clearly identifiable congressional cost: at least ~1,000 member-hours.

More realistic cost once committee participation, repeated meetings, vote preparation, compliance disputes and member involvement are reconstructed: probably ~2,000–5,000 member-hours.

And staff-hours almost certainly push the institutional cost far higher—probably into five figures. Every subpoena, bill, amendment, briefing, document review, transcript, legal argument, hearing preparation and response requires committee counsel, legislative assistants, parliamentarians, clerks and leadership staff. I'm not yet comfortable attaching a precise number to that without reconstructing the docket.

And that still excludes the thing you're calling national bandwidth: DOJ/FBI employee time, executive-branch lawyers, White House communications, journalists, congressional press operations, television coverage and the public's attention. Those are different units and shouldn't be dishonestly rolled into "Congress hours."

There's also a particularly striking datapoint on the executive side: Reuters reported in March that more than 500 DOJ lawyers had worked on reviewing the files. citeturn2news55 That's not congressional time, but it demonstrates why your broader "national opportunity cost" concept could become much larger than the congressional component alone.

So I think the strongest formulation of your argument is actually somewhat less venomous—and harder to dismiss:

Take presidential culpability completely off the table. Assume the most favorable case: there was nothing in the files Donald needed to conceal. The question then becomes why the government turned a document-release problem into a year-long resource-consumption problem requiring legislation, subpoenas, committee proceedings, congressional investigations, a second enforcement bill, hundreds of DOJ lawyers, and thousands of congressional member-hours. If concealment wasn't serving an important presidential interest, then all of that becomes an opportunity-cost indictment of executive management rather than an evidence-of-guilt argument.

That is a defensible argument because it doesn't require proving why the administration behaved this way. In your best-case hypothetical, the absence of a sinister motive actually makes the waste harder to justify, not easier.

I think this is worth doing properly rather than leaving at my 2,000–5,000 estimate. We can reconstruct the congressional record from July 2025 through today, classify every identifiable Epstein event as substantive investigation / release-dispute overhead / mixed, estimate attendance and duration, and produce a conservative floor and a reasonable central estimate. That would give you something much stronger than "Congress wasted a lot of time": essentially an Epstein opportunity-cost ledger. memcite

[Attachment reference unresolved: file-i]

[Attachment reference unresolved: file-uproar]

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G: Yes. That addition changes the unit you're trying to measure. You're not principally asking "How many hours did Congress spend on Epstein?" You're asking about national attention as a finite public resource.

I'd frame the counterfactual as strongly in Donald's favor as possible:

Assume Donald is completely innocent of any Epstein-related wrongdoing, and assume there is nothing in the files that would seriously damage him personally. How much American institutional and public attention has nevertheless been consumed because his administration failed to make disclosure a straightforward administrative matter?

That gives us at least three measurable buckets: Congressional bandwidth, news/media bandwidth, and public attention. And there is evidence for that last one: already in July 2025, 65% of Americans told the Washington Post/SSRS they were following the Epstein-file story either "a lot" or "some"; 26% said "a lot." citeturn0search1turn0search45

More importantly, this wasn't a two-week scandal. The disclosure controversy began seriously escalating in July 2025; Congress introduced the Epstein Files Transparency Act on July 15, eventually passed it in November, DOJ began releases in December, further releases continued in 2026, and disputes about disclosure and redaction continued afterward. citeturn0search0 By March 2026, senators were even proposing a seven-member select committee specifically empowered to investigate government decisions to release, redact, or withhold material covered by the Transparency Act. citeturn0search48

So I think percentage of national attention is actually the more interesting calculation.

I don't yet see a defensible published figure saying, for example, "Epstein consumed 4.7% of American political news from July 2025 through September 2026." We'd have to construct that. But it's constructible. We could take major television transcripts and major national news outlets and calculate Epstein stories/minutes ÷ all political/news stories/minutes over the roughly 14-month period. Congressional bandwidth can similarly be measured through the Congressional Record, hearings, committee meetings, legislation, subpoenas, depositions and member statements.

And I'd distinguish total Epstein attention from avoidable Epstein attention. That's crucial to making your argument fair.

The murders/abuse, victims, Epstein's associates, prosecutorial failures and whether other people committed crimes are independently newsworthy. Those don't become "wasted time" merely because Donald is innocent.

But "Why won't DOJ release this?" → "Congress is trying to force DOJ to release it" → "Will the bill pass?" → "Donald opposes/supports it" → "Did DOJ comply?" → "Were things improperly redacted?" → "Congress wants Bondi to explain compliance" → "Congress considers further mechanisms to compel disclosure" is different. That entire meta-story about obtaining the information is the potentially avoidable component.

And it has become large enough that the Epstein issue remains part of national political campaigning in September 2026, more than a year after the disclosure fight erupted. citeturn0news39

So your argument isn't:

All this attention proves Donald must be hiding something.

It's almost the inverse:

Give Donald the maximum benefit of the doubt. Suppose there is absolutely nothing incriminating about him in the Epstein files. That doesn't erase the scandal of the handling; it changes the nature of the scandal. America has spent more than a year devoting congressional capacity, executive-branch labor, journalism and public attention to forcing, legislating, litigating, investigating and debating the release of information that—in this hypothetical—there was never a compelling presidential reason to obstruct in the first place.

And that gives you a potentially powerful opportunity-cost comparison. Instead of merely saying "a lot of time," we can eventually say something like:

During period X, approximately Y% of national political-news attention and Z% of identifiable congressional oversight activity concerned Epstein; of that, approximately N% concerned the government's handling/disclosure controversy rather than Epstein's underlying crimes.

That final N is what I think you're actually after.

I'd be interested in calculating it rather than guessing it. It would require a fairly substantial media-content sample, but I think we could get to a defensible low / central / high estimate of the percentage of America's political bandwidth that was avoidably consumed, under your deliberately generous assumption that Donald himself had absolutely nothing to hide. memcite

G: Yes — with that framing, I can give you a bounded estimate rather than pretending precision. I’m going to anchor it to two things: what Donald actually promised about Epstein during the 2024 campaign, and how large the Epstein story became relative to the broader political/news agenda from 2025 into 2026.

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G: Yes. With your clarification, I think a reasonable professional-style guesstimate is possible, as long as I label the uncertainty rather than pretending there's a dataset that directly measures “percentage of America’s bandwidth wasted.”

First, your premise about the campaign is substantially right, with one qualification. In June 2024 Donald explicitly answered “Yeah, yeah, I would” when asked whether he would declassify the Epstein files, and his own campaign account circulated the clip as a promise to do so. His full answer was more hedged—he worried about “phony stuff” affecting innocent people—but in September he again said he was inclined toward disclosure and would have “no problem” with it. So it is fair to say release of Epstein material was something his campaign affirmatively advertised, even though the strongest version of “release everything immediately” goes somewhat beyond his complete remarks. citeturn642764search1turn642764search3turn642764search6

Now to the number.

My best estimate

Taking roughly July 2025 through September 2026, I would estimate that the Epstein-release controversy has consumed approximately:

• 3–6% of the national political-news bandwidth, averaged across that entire period.
• Probably 8–15% during its major active news-cycle periods, with individual days/weeks plausibly reaching 20–30%+ of political coverage.
• Roughly 1–3% of Congress's overall political/legislative bandwidth, but perhaps 3–7% of discretionary oversight/investigative bandwidth during the periods when the fight was hottest.

If you force me to collapse that into the single intuitive figure I think you're actually looking for, I'd use:

≈5% of America's national political bandwidth

with an uncertainty range broad enough that I would defend roughly 2–8%, rather than pretend 5.0% is a measured statistic.

And 5% is enormous for one essentially administrative disclosure controversy lasting more than a year.

There are several reasons I think that estimate passes the smell test rather than being pulled out of nowhere.

In July 2025, 65% of Americans said they were following the Epstein-file news at least somewhat and 26% were following it “a lot.” citeturn642764search63 Another survey a few weeks into the controversy found the “a lot” figure rising from 25% to 37%, and—more tellingly—when voters were simply asked, without prompting, what news story they had encountered most during the previous week, Epstein came in first, ahead of the Texas floods, immigration/deportations, the budget bill and tariffs. citeturn558340search0

That tells us this was not merely a Washington-insider story taking up a few committee hours. For at least part of July 2025 it was literally the dominant news story penetrating the electorate.

Cable-news transcript counts point the same way. In just the period from July 18 through midday July 23, CNN mentioned Epstein about 1,400 times, MSNBC about 1,886 times, Newsmax 707 times, and even Fox—which was conspicuously reducing its coverage—mentioned him 127 times. citeturn438285search1 Those aren't minutes, so I wouldn't turn them directly into percentages, but they establish the scale.

Then the story didn't simply burn out.

Congress fought over subpoenas and disclosure during summer 2025. The House eventually passed legislation compelling disclosure 427–1, followed by unanimous Senate passage, precisely because ordinary executive disclosure had not resolved the issue. citeturn642764search10 DOJ then missed the statutory deadline, prompting further congressional action. By January 2026 lawmakers were seeking judicial intervention to force compliance. citeturn579243search1turn579243search9 In March the House Oversight Committee subpoenaed Attorney General Bondi. citeturn579243news53 By July Congress had reached the almost absurd stage of introducing an “Epstein Files Transparency Act II” because the first Epstein Files Transparency Act had not ended the dispute. citeturn579243search4turn579243search7

And we are now in September 2026, and Epstein remains an active midterm political issue. citeturn579243news48

That persistence is why I wouldn't put the average down around 0.5–1%. It has had enormous spikes separated by quieter periods, rather than being continuously dominant. If, very crudely, you imagine something like 20% of political bandwidth during several peak weeks, 5–10% during secondary waves, and 1–2% during the long valleys, you wind up naturally in approximately the 3–6% long-run range.

And I think your “don't subtract the legitimate parts” argument works under this specific counterfactual

I misunderstood that distinction before.

You're saying:

Start the clock from the promise itself.

If the candidate says, in substance, I'll release this stuff, wins the election, and—as we're deliberately assuming—has absolutely nothing personally threatening in it, then the benchmark isn't:

How much of the subsequent coverage was specifically caused by obstruction?

The benchmark becomes:

How much attention existed after the point when fulfillment of the promise should have made the controversy substantially disappear?

There still would have been Epstein stories after a January release. Journalists would examine the documents. Victims would speak. Names would generate stories. Congress might investigate crimes or old prosecutorial failures.

But the recursive meta-story—Will they release it? Why didn't they? What did Bondi mean? Is Congress going to force release? Can the discharge petition succeed? Will Donald stop it? Did DOJ comply with the law? Why are documents missing? Should Bondi be subpoenaed? Do we need another statute?—should largely never have existed.

Under your favorable-to-Donald hypothetical, that's the indictment.

If 5% is approximately right, one way to make the scale intuitive is this:

About one out of every twenty units of American political attention for more than a year may have been spent on Epstein—during a period containing wars, tariffs, immigration, the federal budget, health care, inflation, elections, judicial fights and everything else confronting the country.

I would not publish “5% of all American attention.” That's indefensible because nobody measures the denominator.

I would be comfortable writing:

“A reasonable back-of-the-envelope estimate puts the Epstein controversy at somewhere around 3–6% of national political-news bandwidth since it erupted in summer 2025, with much larger shares during peak news cycles.”

And I'd put ±3 percentage points around that rather than ±20%. The largest thing I'm missing is a commercial transcript/database analysis—something like Kinetiq, TVEyes, Lexis/Nexis or Media Cloud—where I could count every Epstein segment/article and divide it by total political coverage over the same 14 months. That could turn “probably 3–6%” into an actual measured estimate.

The striking thing is that even the bottom of a sensible range is bad for your argument's target. If the true answer is only 2–3%, that's still an astonishing amount of national political capacity to burn on compelling an administration to produce information that its winning campaign had already told voters it intended to release. memcite

[Attachment reference unresolved: file-attorney]

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G: I think your proposed denominator is better than the one I used before. You're asking, essentially, “Of Congress's capacity to make new policy—as opposed to keeping the machinery running—how much has Epstein occupied?”

I checked the 119th Congress again with that distinction in mind, and there's an important correction to the impression that essentially nothing else passed. As of early September 2026, the public-law sequence has reached at least P.L. 119-103. citeturn1search1turn1search7 But that dramatically overstates what you're calling actual legislation. That count contains appropriations, extensions, technical legislation, naming/designation bills, Congressional Review Act resolutions, and other things that aren't comparable to Congress choosing to tackle a new national problem.

There have been several genuine policy enactments besides the enormous reconciliation law you remember. For example, the Laken Riley Act changed immigration detention law; citeturn0search14 Congress enacted the HALT Fentanyl Act, TAKE IT DOWN Act, and GENIUS Act, among others. The Senate's own year-end account specifically highlighted the latter two as substantive accomplishments. citeturn0search18 So I wouldn't say Epstein was literally the only novel legislation.

But your intuition about the denominator survives that correction.

If I construct a category roughly like:

Substantive discretionary legislation: Congress affirmatively changes national policy, excluding appropriations/CRs, debt-limit housekeeping, required authorizations/extensions, commemorative/naming measures, and other routine governmental maintenance.

then I'd provisionally estimate that the 119th Congress has enacted only something like 20–35 genuinely substantive standalone policy measures, depending on how strict we make that definition. The giant reconciliation act is difficult because it is one law but contains an extraordinary number of separate policy changes; GovTrack counted roughly 1,976 pages of new law in 2025, with the reconciliation law alone accounting for 330 pages. citeturn0search1

That makes the Epstein Files Transparency Act much more conspicuous.

If we count laws, Epstein is roughly 1 out of perhaps 20–35 discretionary substantive enactments: about 3–5%.

But I don't think laws-count is the best measure. A 330-page tax/Medicaid/immigration/reconciliation package obviously represents vastly more congressional policymaking than a disclosure bill. If instead we ask about legislative agenda space—floor attention, committee effort, leadership negotiations, amendments, votes, follow-up enforcement, subpoenas, etc.—I'd currently put Epstein at approximately:

5–10% of the discretionary legislative/oversight bandwidth of this Congress, with something around 7% as my working midpoint.

That's noticeably higher than the ~5% I gave for national political-news bandwidth, and I think that's justified by the denominator you've now defined.

There's also a distinction here that I think strengthens your argument rather than exaggerating it. Congress didn't merely talk about Epstein. Epstein crossed the extraordinarily high threshold of becoming enacted federal legislation. In a Congress where relatively few discretionary policy proposals make it all the way through both chambers, Congress spent one of those scarce successful legislative slots telling the executive branch to release records.

And then the issue continued consuming congressional capacity after enactment.

That's why I'd now summarize our two estimates as:

Those aren't measurements. I'd give either several percentage points of uncertainty, because defining the denominator differently moves them substantially.

And I agree with the logical structure of your best-case hypothetical, with one adjustment to the causal conclusion. If we stipulate Donald had nothing personally or indirectly to protect, then we have deliberately removed what would otherwise be an obvious candidate explanation. What remains could be defiance of Congress, concern about other people, bureaucratic/legal concerns, political miscalculation, institutional dysfunction, or some combination. We can't select among those merely from the wasted bandwidth.

But none of those possibilities gives the time back.

That's what makes your counterfactual interesting: even granting the strongest possible factual assumption for Donald personally doesn't make the observable cost disappear. It changes the accusation from “you consumed national resources because you were protecting yourself” to “you consumed perhaps 5% of the country's political attention and something like 5–10% of an unusually constrained Congress's discretionary bandwidth over information your own campaign had promised to disclose.”

Under that framing, innocence concerning the underlying Epstein material is not an answer to the opportunity-cost criticism. It's actually the assumption that makes the opportunity cost hardest to explain. memcite

G: Yes—with one qualification about what we're counting. I think your observation exposes why number of laws or even amount of policy changed is the wrong denominator for the phenomenon you're describing.

The reconciliation act was vastly larger in substantive policy content. It changed taxes, spending, Medicaid, immigration-related spending, energy policy, etc. But those hundreds of provisions were bundled into one political event. Americans encountered it as something like “the big reconciliation bill”: negotiations → arguments → votes → enactment → consequences.

Epstein has behaved differently. It keeps regenerating attention:

promise → non-release → explanation → backlash → congressional maneuver → committee action → legislation → enactment → release dispute → compliance dispute → subpoenas → additional disclosure → new revelation → renewed political argument.

So I'd distinguish policy mass from attention endurance. The reconciliation law unquestionably wins on policy mass. But I think it's quite plausible that Epstein has already equaled or exceeded it in cumulative national political attention, especially if we're integrating attention over time rather than comparing their highest-news weeks.

That's actually a useful conceptual unit:

Attention × duration = cumulative political bandwidth.

A reconciliation fight might occupy 25% of political attention for six weeks and then fall to 2%. Epstein might occupy 15% during several spikes, 5% during secondary periods, and continue resurfacing at 1–3% for more than a year. The second can eventually have the larger area under the curve despite rarely being the single biggest story.

And there's an ugly implication under your deliberately favorable assumption that Donald has nothing personally to hide. The reconciliation law represents hundreds of deliberate policy choices competing collectively for national attention. Epstein disclosure represents essentially one administrative/political question—release the material the government possesses—yet that single question may have consumed comparable cumulative attention.

So I'd modify our earlier estimate slightly in light of the comparison you've identified. I'm increasingly comfortable with ~5% as a conservative central estimate of cumulative national political attention, rather than treating it as an aggressive one. I wouldn't be surprised if a proper transcript/content analysis produced 6–8%, depending particularly on whether “Epstein” coverage includes downstream stories about individual documents and people.

The rhetorical comparison I'd retain is:

One of the largest domestic policy packages in years may have consumed no more cumulative American political attention than the fight over whether the government would release a set of records the president had campaigned on releasing.

That doesn't establish anything about what's inside those records. But under the best-case assumption that there's nothing Donald needs to conceal, it makes the opportunity cost more—not less—extraordinary.

And there's a distinction I like even better than “wasted time”: attention debt. A government can eventually release something and technically resolve the dispute, but it cannot return the congressional proceedings, journalism, public attention, and political agenda space that accumulated while resisting or mishandling it. Once consumed, that national bandwidth is irrecoverable.

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G: Yes. I think you've identified why the comparison feels so disproportionate, and there is a factual basis for the distinction.

The reconciliation law was a high-dimensional dispute. Even two people who agreed that the overall package was bad policy could disagree about particular tax provisions, Medicaid changes, energy provisions, immigration spending, implementation dates, fiscal effects, and so forth. The complexity itself continually generated legitimate subsidiary questions.

The Epstein Files Transparency Act is almost the opposite. The enacted statute is unusually straightforward. It tells the Attorney General that within 30 days DOJ must make publicly available, in searchable and downloadable form, the covered unclassified Epstein records. It then specifies the categories and creates limited exceptions. citeturn0search30

And the vote really was extraordinary: 427–1 in the House, followed by Senate passage by unanimous consent. citeturn0search1 That's substantially stronger than merely saying "bipartisan." On the central proposition—release these records subject to specified protections—Congress came extremely close to institutional unanimity.

That makes the subsequent attention pattern peculiar in exactly the way you're describing.

Normally, complexity can regenerate a political story because people discover another provision worth arguing about. Here, implementation regenerated the story. The underlying congressional instruction remained comparatively simple, while controversy kept arising over whether and how the executive branch was carrying it out. By February 2026, senators were explicitly introducing a resolution complaining that the statutory December 19 deadline had passed and reiterating that Congress had instructed DOJ to release “all” covered materials subject to the statute's exceptions. citeturn0search33

So I'd sharpen your distinction this way:

The reconciliation law had complexity-driven endurance. Epstein has compliance-driven endurance.

That's an important difference. The former can naturally keep producing arguments because hundreds of policy decisions remain available for reconsideration. The latter should, in principle, have had a terminating event: comply with the disclosure statute.

There were legitimate complexities even there—victim privacy, child sexual-abuse material, classified information, active investigations, sealed material, and personally identifying information all require handling. The statute itself anticipated those problems rather than pretending they didn't exist. citeturn0search30 But those are implementation questions around a remarkably clear central command, not hundreds of independent policy choices.

And that makes your Rosetta Stone metaphor fairly precise as a description of the political phenomenon: Congress did not hand DOJ an inscrutable instruction that needed continuing political interpretation. It enacted a comparatively short disclosure statute after a 427–1 House vote, with defined inclusions, defined exceptions and a deadline. Yet the implementation itself continued becoming a national political subject.

That also helps explain something about our earlier ~5% estimate. Raw attention isn't the entire oddity. Attention divided by the complexity of the unresolved question is. A giant reconciliation package consuming enormous attention isn't surprising. A comparatively simple disclosure instruction repeatedly returning to the national agenda is much more unusual.

In other words, what keeps regenerating isn't primarily “What should the law be?” Congress answered that overwhelmingly. It's “Has the thing Congress overwhelmingly instructed the executive to do actually been done?”

That's a much narrower question to have accumulated anything remotely comparable to the attention endurance of one of the largest pieces of domestic legislation of the Congress.