NFL revenue sharing explained: the deal that made the league

A publicly owned non-profit corporation in a city of about a hundred thousand people competes on equal financial terms with clubs in New York, Los Angeles and Dallas. Not approximately equal. Equal, in the specific sense that the two receive identical distributions from the same pot and face an identical ceiling on what they may spend on players. NFL revenue sharing is the arrangement that does that, and it is usually explained as "the teams split the television money", which is true and stops about four layers short of the interesting part.

Nothing else in professional sport looks like it. The Premier League shares its central money on a formula deliberately tilted towards finishing position and appearances. European football's biggest clubs have spent thirty years negotiating a larger slice for themselves. Baseball has no ceiling at all. The NFL took the opposite decision in the 1960s and has never seriously revisited it, and the whole competitive character of the sport follows from that choice.

What sits underneath the summary is three separate mechanisms stacked on each other, a written definition of "revenue" running to thousands of words, a set of weights that value a pound of local money differently from a pound of national money, and a clause that lets an owner reduce the players' share by building a stadium.

All of it is public. Most of it is in one article of one agreement, and almost nobody reads it.

Three systems, one phrase

Before anything else, separate the mechanisms, because they are routinely blurred together and they answer different questions.

Sharing between clubs. The league collects money centrally and distributes it. This is what people usually mean, and it is governed by the league's own constitution and bylaws rather than by any agreement with the players.

Sharing between owners and players. The collective bargaining agreement takes a defined slice of everything and hands it to the players, and it is that slice, divided by 32, that becomes the salary cap. This is governed by Article 12 of the agreement and is where all the precise language lives.

Sharing of gate receipts. A distinct, older mechanism by which the money taken at the turnstile is not entirely the home club's to keep. The players' agreement refers repeatedly to "gate receipts subject to revenue sharing", which tells you the pooling exists without the agreement itself setting the formula.

The first determines whether small-market clubs can compete. The second determines what the players earn. The third is the residue of an older world in which the gate was the business. Confusing them produces most of the bad analysis on the subject, including the recurring claim that a club "cannot afford" a signing, which under a hard ceiling funded from an equal distribution is close to meaningless.

Why the national pot exists at all

The pooled sale of broadcast rights was not obviously legal. Thirty-two separate businesses agreeing to sell their television rights jointly, through a single agent, at a single price, is the kind of arrangement antitrust law exists to prevent, and a court said so.

Congress responded in 1961 with legislation specifically permitting professional sports leagues to pool and sell their television rights collectively. That statute is the foundation everything else sits on. Without it, each club would negotiate its own broadcast deal, the clubs in the largest markets would command the largest fees, and the sport would have developed the shape European football has: a handful of financially dominant clubs and a long tail that exists to play them.

Instead the league sells the packages centrally and divides the proceeds equally. A club's television income does not depend on how many people live near it, how often it appears in prime time, or whether it won a single game. The distribution is the same for the champion and for the club that finished bottom.

That is a genuinely radical arrangement and it has an obvious cost to the largest clubs, who could plainly earn more alone. They accept it for a reason that has held for six decades: an equal league is a more valuable league, because uncertainty about outcomes is what people pay to watch, and a competition in which four clubs can win is worth less in aggregate than one in which any of them can. The owners of the largest clubs are not being generous. They are taking a smaller share of a much larger number.

The contrast with how the Premier League distributes its central money is instructive precisely because that system also shares, and shares a lot, but keeps a merit component and a facility-fee component that reward the clubs already winning. Two leagues, two philosophies, and a competitive spread that reflects them.

What counts as revenue, according to the people who argue about it

Here is where the subject stops being a summary and becomes a document.

Article 12 defines "All Revenues", abbreviated to AR, as the aggregate revenues received by the league and all its clubs, from all sources, whether known or unknown, derived from, relating to or arising out of the performance of players in NFL football games. Every word of that is doing work. "All sources" and "whether known or unknown" exist so that a revenue stream invented in 2027 is captured without renegotiation. "Arising out of the performance of players" is the test that decides the hard cases.

The agreement then lists what is included, and the list is more specific than any summary suggests.

Gate receipts from the preseason, the regular season and the postseason, including money from luxury boxes, suites and premium seating, counted net of admission taxes and of surcharges paid to stadium or municipal authorities. For a premium seat, the amount that counts is the face value of the ticket. A service fee charged on a ticket account rather than per ticket, up to a maximum set by league policy, is not revenue. Credit card charges are explicitly not treated as a deductible surcharge, which closes an obvious route to shrinking the number.

Broadcast money of every description: network, local, cable, pay television, satellite, international, delayed, and any other means of distribution, along with copyright tribunal and extended market payments.

Concessions, parking, local advertising, signage, magazine advertising, local sponsorship, stadium clubs, internet operations including merchandise sales, and sales of programmes and novelties.

The consolidated revenue of NFL Ventures, the partnership through which the league runs its properties, its films operation and its media enterprises.

Barter income, valued at ninety per cent of the fair market value of whatever was received. Equity instruments received from third parties, brought in at fair value on vesting and amortised over ten years, with the carrying amount adjusted using an option pricing model in the intervening years.

Revenue from a stadium lease with an unrelated third party where the activity has nothing to do with football, provided the club did not have to make a meaningful investment of capital to earn it. Recoveries under business interruption insurance, net of premiums and deductibles, to the extent they replace revenue that would have counted. Expense reimbursements from government bodies connected to a stadium lease.

And gambling revenue, which is drafted with a precision that repays reading.

1. 1The ticket is soldFace value counts. Admission taxes and municipal surcharges come off first. A per-account service fee under the league policy cap does not count at all, but credit card charges cannot be deducted.
2. 2It lands in a revenue bucketGate money is Local revenue, because it is received by the club. Broadcast rights fees are League Media. Anything from the league's own trading arms is Ventures and Postseason. No dollar may sit in two buckets.
3. 3Everything is summed into All RevenueThe league and the union both appoint accountants, who report the figure. Projected revenue for the coming year is what the calculation actually runs on, with a later true-up against the real number.
4. 4Bucket percentages are appliedFifty-five per cent of League Media, forty-five per cent of Ventures and Postseason, forty per cent of Local. The sum of those three is the starting player cost amount.
5. 5The bands clamp itIf the result exceeds 48.5 per cent of projected revenue it is cut back to 48.5. If it falls below 48 per cent it is raised to 48. The players' share is a corridor, not a percentage.
6. 6The stadium credit is deductedApproved private stadium spending reduces the player cost amount, but never below the floor. This is the one lever an owner can pull that moves the players' side of the ledger.
7. 7The cap is what is left, divided by 32Benefits come out, the remainder becomes the salary cap, and every club receives the identical number regardless of what it earned.

The route every dollar of football revenue takes through the agreement. The percentages at step four are the ones written into Article 12.

The gambling clause, and what careful drafting looks like

Gambling revenue is included in full, and the agreement spends several hundred words defining which gambling revenue.

A betting operation physically inside a stadium, or attached to it, is treated as being "in" the stadium. One within two hundred feet of the outer wall, or of the point at which a ticket is required, whichever distance is greater, is "near" it. Everything received during the season, defined as running from the start of training camp to the club's last game, counts in full. Out of season, only half of the revenue from an operation inside the stadium counts, and only a third of the revenue from one near it.

If the club merely holds a minority stake in the betting business rather than receiving the revenue, the money counts pro rata to that stake, under the same geographic and seasonal formula. If the stake is five per cent or less and passive, none of it counts.

Read that back and you can reconstruct the entire negotiation. The union wanted gambling money captured, because it is plainly derived from football. The owners wanted their non-football property development kept out. What emerged was a rule measured in feet and calendar dates, which is what happens when two sides with good lawyers agree that a principle is right and cannot agree where it stops.

The slot machines matter here. The clause explicitly captures revenue from non-football gambling, a slot machine in a stadium concourse included, because the machine only exists where it is on account of the football. That is the "arising out of the performance of players" test applied at its outer edge, and it is the single clearest illustration of how wide the definition is meant to run.

The three buckets, and why a local dollar is worth less

This is the part most explanations miss entirely, and it is the mechanical heart of the system.

All revenue is subdivided into three categories for the purpose of calculating what the players get.

League Media is money from selling the right to show entire games: national and regionally packaged television, cable, satellite, internet and other media, international television rights, national radio, and the copyright tribunal. The agreement goes as far as naming the specific broadcast packages that fall into it as of the year it was signed, which removes any argument about where a given contract sits.

NFL Ventures and Postseason is money from the league's own operations: its network, its properties arm, its films business, its digital products, and postseason revenue received by the league rather than by a club.

Local is everything else received by the clubs. The agreement defines it by subtraction, which is the only sensible way to draft it, and specifies that preseason television rights sold by a club belong here.

Now the weights. The starting player cost amount is fifty-five per cent of projected League Media revenue, plus forty-five per cent of projected Ventures and Postseason revenue, plus forty per cent of projected Local revenue.

League Media revenue55%

NFL Ventures and Postseason revenue45%

Local revenue40%

The percentages written into Article 12 for the initial calculation of the player cost amount, before the bands and the stadium credit are applied.

Show the numbers

What each revenue bucket contributes to the players' side

Item	Value
League Media revenue	55%
NFL Ventures and Postseason revenue	45%
Local revenue	40%

A dollar of national broadcast money puts fifty-five cents into the players' pot. A dollar earned at a concession stand puts forty. That gap is not arbitrary and it is not an accident of negotiation. Local revenue carries local costs: staff, operations, the physical business of running a stadium. National media money arrives with almost no cost of production attached. Weighting them identically would have charged the players' account for the clubs' operating expenses, so the parties discounted local money instead of trying to audit thirty-two sets of expenses.

Two protections sit on top of the buckets.

The no migration rule states that no revenue may appear in more than one category, that all revenue must appear in one, and that revenue for substantially similar rights may not move between buckets in later years regardless of which entity receives it. Without that clause, a league that wanted to shrink the players' share could simply route a broadcast contract through a subsidiary and reclassify fifty-five cent money as forty cent money.

The bundling clause handles the case the rule cannot: a future deal that genuinely mixes media rights with other rights. The parties must discuss the allocation in good faith, and if they fail, an allocation arbitrator decides. Each side proposes a split and the arbitrator picks one of the two proposals outright, with no power to invent a middle. That is baseball-style arbitration, and it is chosen deliberately, because a decision-maker who can only pick an extreme forces both sides to propose something reasonable.

What is deliberately not revenue

The exclusions tell you as much as the inclusions, because each one marks a place where the parties agreed that money exists but does not arise from playing football.

Proceeds from selling a franchise, or a stake in one, are excluded. So are expansion fees for granting a new club, and relocation fees for moving an existing one. So are dues and capital contributions paid by clubs to the league, fines, and interest income. So are sales of real estate.

The reasoning is consistent. When an owner sells a club for a very large sum, that sum reflects the accumulated value of a franchise in a league the players built, and the players receive none of it. The justification is that a sale is a transfer of an asset between owners rather than income generated by the performance of players. Whether that is fair is a genuine argument, and it is the argument that will be had at the next negotiation, because franchise valuations have risen faster than revenue for years and the players' share of that appreciation is precisely zero.

Revenue from stadium events unrelated to football, concerts and other sports among them, is excluded where the club had to invest real capital or effort to generate it. Cheerleader revenue is excluded, with a caveat that the accountants will look at it if the counterparty has another commercial relationship with the club.

And there is one exclusion that is almost a joke about the subject of this article: revenue sharing among the clubs is itself excluded from revenue. Money moving from one club to another is not new money, and counting it would inflate the pot with the same dollars twice.

Certain personal seat licence proceeds are also carved out, along with wholesale merchandising conducted by one particular club's own merchandising arm, and those exclusions are the fingerprints of specific historical negotiations rather than principles.

The stadium credit: the lever that moves the players' side

Now the mechanism that almost nobody outside the negotiation understands, and the one that connects this subject to every argument about publicly funded stadiums.

For each league-approved stadium project, there is a credit equal to fifty per cent of the private cost of building or renovating it. In California the figure is seventy-five per cent. Private cost includes financing costs. The credit is amortised over up to fifteen years, beginning in the year before the stadium opens, using a rate based on the league's own long-term borrowing cost.

That credit reduces the player cost amount. In plain terms: when an owner puts his own money into a stadium, half of it comes back out of the players' share, spread over fifteen years.

There is more. The credit also includes seventy per cent of certain personal seat licence proceeds, premium seat revenue and naming rights money that were excluded from revenue in the first place, along with half the cost of capital spending on the fan experience in an existing stadium.

The protection for the players is the floor. However large the stadium credit, the player cost amount cannot be pushed below the agreed percentage of projected revenue, which for the seasons running through the end of the agreement is forty-eight per cent. The credit can take the players from the top of the band to the bottom of it. It cannot take them below.

This clause explains a great deal of behaviour that otherwise looks irrational. It gives owners a reason to spend privately rather than to hold out for the last dollar of public money, because private spending is partly reimbursed by the players. It sets a higher rate in one state, which is a negotiated response to a specific project. And it means that every argument about whether a city should fund a stadium is taking place inside a system where a portion of the private alternative is already subsidised from inside the sport. The broader question of who actually pays for these buildings is worth taking on its own terms, and the economics of stadium finance run well beyond football.

• 48Floor for the players' share of projected revenue
• 48.5Ceiling for the players' share of projected revenue
• 90Minimum team cash spending across each multi-year block
• 50Stadium credit as a share of private construction cost

Figures set out in Article 12 for the League Years running to the end of the current agreement.

Why the players' share is a corridor rather than a number

Add the three weighted buckets together and the result will not naturally land on a round number. It moves with the mix: a year in which national media grows faster than local revenue pushes the sum up, because more of the total is being counted at fifty-five per cent instead of forty.

The agreement handles that with bands. If the calculated amount comes out above the ceiling, it is cut back to the ceiling. If it comes out below the floor, it is raised to the floor. For the seasons running to the end of the agreement, that corridor is half a percentage point wide.

The consequence is worth stating plainly, because it changes how the whole system should be read. The bucket percentages do not really determine the players' share. The bands do. The percentages determine where inside the corridor the number lands, and whether the stadium credit has anything to bite on. Almost every year the calculation runs up against one edge of the band or the other, which is why the headline share barely moves while the underlying revenue mix changes constantly.

That is a deliberate design. Both sides preferred a predictable share to an accurate one. Predictability is what allows a club to plan four years ahead and what allows the union to know roughly what it is bargaining over, and the loss of precision is a price both were willing to pay. The knock-on effects for how clubs actually build rosters against that ceiling belong with the cap itself, which is the same money viewed one step further downstream.

The media kicker, and what a seventeenth game was worth

The agreement contains a mechanism that priced, in advance, something that had not happened yet: the possibility that the next round of broadcast contracts would be far larger than the last.

The parties agreed a baseline figure for what national media had been worth. Averaged across the 2014 to 2022 seasons, using a defined list of packages, they agreed that the current average was 7.357 billion dollars a year. That figure is written into the document as an agreed fact rather than an estimate, which is how you avoid arguing about the past.

They then set a threshold. If new media contracts, negotiated after the agreement and covering a seventeen-game season, achieved an average annual value more than thirty-five per cent above that baseline, the players' share would rise. Thirty-five per cent above 7.357 billion is 9.932 billion, and the agreement states that arithmetic explicitly.

Agreed baseline, 2014 to 2022 average7.36bn

Kicker threshold, 135 per cent of baseline9.93bn

The agreed baseline for national media revenue and the level new contracts had to beat for the players' share to rise. Both figures are stated in Article 12.

Show the numbers

The media kicker threshold, as written into the agreement

Item	Value
Agreed baseline, 2014 to 2022 average	7.36bn
Kicker threshold, 135 per cent of baseline	9.93bn

Above the threshold, the increase is slotted. The actual percentage by which the new deals beat the old ones is calculated, rounded to a hundredth, and looked up against a schedule appended to the agreement, which returns the additional share of revenue the players receive. The schedule is illustrated in the text: a sixty per cent increase takes the players to the top of their band, and beyond that the league is entitled to recoup part of the gain.

This is the clearest example in the whole agreement of both sides pricing an unknown. The players were being asked to add a game to the season, with everything that implies for careers and bodies. Rather than accept a fixed payment, they took a claim on the specific revenue that the extra game was expected to generate, structured so that they gained only if the broadcast market delivered the increase the owners were forecasting. If the market had disappointed, the kicker simply would not have applied.

The floor: clubs cannot pocket the shared money

An equal distribution creates an obvious temptation. A club that receives the same national money as everyone else, and spends less of it on players, converts the difference into profit. The salary cap alone does nothing to prevent that, because a ceiling is not a requirement.

So the agreement adds a floor, and it is a cash floor rather than a cap floor, which matters. The two are different numbers, because cap accounting spreads signing bonuses across years while cash accounting records them when paid.

The requirement is expressed across multi-year blocks rather than single seasons. Across each defined block of seasons, every club must spend a minimum percentage of the salary caps for that period in actual cash. A club that comes up short pays the shortfall directly to the players who were on its roster at any point during those seasons, allocated as the union directs, by a fixed date after the block closes. If it fails to pay, the league pays on its behalf.

Two features of that design are worth noticing. The multi-year window is what lets a club underspend deliberately in a rebuilding season and make it up later, which is a legitimate strategy the parties chose to permit. And the remedy is a direct payment to the affected players rather than a fine to the league, which removes any incentive for the league to be relaxed about enforcement.

What the system does not equalise

Equal distributions and a hard ceiling produce a competitive balance nothing else in major professional sport matches. They do not produce equality, and the places where the system leaks are where the remaining advantage sits.

Local revenue is not shared. A club with a new stadium, a large corporate base and a deep local sponsorship market keeps that money. It cannot spend it on players beyond the cap, but it can spend it on everything the cap does not cover: coaching staff, scouting departments, sports science, facilities, analytics, and the sheer number of people employed to make marginal decisions slightly better.

Cash is not capped the way the cap is. Every club faces the same ceiling on charges, and no club faces a ceiling on cash paid out in a given year. An owner willing to fund a large cash outlay can structure contracts in whatever way suits the football decision. One who is not must structure them around cash flow. Over a decade that compounds quietly.

Stadium subsidy is unequal. A club that persuaded a public authority to fund most of its building has a different cost base from one that borrowed to fund its own, and the stadium credit only partly offsets the difference.

Ownership wealth is unequal and unregulated. Nothing in the agreement addresses it, and the exclusion of franchise sale proceeds from revenue means the largest single source of owner enrichment sits entirely outside the shared pot.

The honest summary is that the NFL has equalised the input that most directly determines who wins on Sunday, and left almost everything else alone. That is a much better result than any other major league has achieved, and it is not the same thing as parity. The alternative model, in which clubs rise and fall on their own commercial performance, is set out in the case for and against promotion and relegation, and the comparison across leagues is stark enough to be worth reading beside this.

Why the Packers are the reason any of this is visible

One club is a publicly owned non-profit corporation with hundreds of thousands of shareholders, and it therefore reports its finances in public every year.

That single accident of history is why anyone outside a boardroom knows what an NFL club's national distribution actually is. Thirty-one clubs are private companies that disclose nothing. One holds an annual meeting, publishes figures, and separates the money that came from the league from the money it earned locally.

Every serious estimate of NFL club finances is calibrated against that one set of accounts, on the reasonable assumption that the national distribution is identical across clubs, which is exactly what the sharing arrangement guarantees. It is a strange way to run an information environment, and it is a useful reminder of how little of this would be checkable if one small Wisconsin corporation had incorporated differently in the 1920s.

What to look at, if you want to judge a club's finances

Four things, and none of them is the figure that gets reported.

The national distribution, not the club's total revenue. Every club receives the same number. A club described as one of the league's richest is being described on its local margin, which is the smaller part of its income.

Local revenue growth, not local revenue level. The level reflects the market. The growth reflects the operation, and it is the only part of the top line a management team genuinely controls.

Cash spending against the floor, over the block rather than the season. A club well below the floor two years into a three-year period has a bill coming, and it will be paid to players rather than to the league.

Stadium spending and where it sat in the credit. Private construction spending reduces the players' share for up to fifteen years. A club that has recently built is carrying a credit, and that credit is one of the few things capable of moving the sport's central number.

Those four are all findable, all stable, and all more informative than a valuation estimate produced by somebody who has never seen the accounts. The system that generates them was designed in the 1960s by owners who understood that their competition was worth more shared than divided, and it has survived every commercial pressure since, which is a reasonable claim to being the most successful piece of sports administration anyone has attempted. More on how the sport's money, contracts and competition rules fit together sits in the American football archive.