Economics
How the NFL salary cap works, and why cap hell is a myth
The NFL salary cap explained properly: where the number comes from, why cash is not cap, how signing bonus proration works, and what dead money really costs.
By CricketTaken EditorialPublished Economics17 min read
Every February, a certain kind of article appears. A club is said to be "in cap hell". The number quoted is enormous, usually eight figures and negative, and the conclusion is always that the team is about to be dismantled. Then March arrives, the club signs three players it was not supposed to be able to afford, and nobody writes the follow-up.
This happens because the salary cap is reported as though it were a bank balance and understood by almost nobody as what it actually is: an accounting system with a set of rules that a competent front office can bend a very long way. The rules are not secret. They are written down in the collective bargaining agreement, they have barely changed in structure since 1994, and once you understand four of them the annual panic stops making sense.
This is those four rules, and what follows from them.
What the cap actually limits
The salary cap is a per-club ceiling on a defined quantity of money in a defined twelve-month period. It is not a limit on what an owner may spend. It is a limit on what a club may account for.
Two things make the NFL version unusual among the major leagues. The first is that it is genuinely hard. Baseball has no cap at all, only a competitive balance tax that a determined owner treats as a cost of doing business. Basketball has a soft cap so riddled with exceptions that most contending teams spend the season above it. Hockey has a hard cap much closer to this one, though it counts contracts differently. The NFL has neither a tax nor a meaningful escape hatch. A club must be under the number when the league year opens in the middle of March, and it must stay under it. There is no cheque anyone can write to be over.
The second is that the contracts underneath it are, for the most part, not guaranteed. In the NBA and MLB, a signed contract is money the player will receive. In the NFL, the guaranteed portion is negotiated separately and is often a fraction of the headline figure. A five-year, $100 million deal in football may mean forty guaranteed and a club option to walk away every March thereafter. This single fact is why cap management in the NFL is a different discipline from cap management anywhere else: the club retains the right to end almost any contract, and the only question is what that costs.
- 5Maximum years a signing bonus may be prorated over
- 51Contracts counted against the cap in the offseason
- 2Post-June 1 release designations allowed per club
- 53Players on an active roster in the regular season
Where the cap number comes from
The cap is not set by a commissioner with a spreadsheet. It is a share of revenue, calculated from a formula the two sides negotiated, and the players' share is close to half of everything the league takes in.
The mechanism runs backwards from that share. Total league revenue for the relevant period is measured, the players' negotiated percentage of it is calculated, adjustments are made for benefits and for the previous year's actual spending, and what remains is divided by the number of clubs. That quotient is the cap. Every club gets the identical figure, which is the part people forget: the richest team in the league and the poorest have exactly the same ceiling.
Two consequences follow, and both matter more than they sound.
Because the cap tracks revenue, it rises when the league's media deals rise. Those deals have risen steeply and predictably for three decades, which means a contract that looks enormous against today's cap will look ordinary against the cap three years from now. Front offices know this and price it in. A deal that consumes a fifth of the cap in year one may consume an eighth by year four without a single dollar changing, simply because the denominator grew.
And because every club has the same number, competitive advantage in the NFL cannot come from spending more. It has to come from spending better, or from spending sooner. That is a genuinely different game from the one played in the Premier League or in European basketball, and it is the reason the NFL's competitive spread is narrower than any comparable competition.
The distinction everything else rests on: cash is not cap
Here is the sentence that unlocks the rest. The money a club pays out in a year and the money it charges to the cap in that year are two different numbers, and they are allowed to be different.
A player's cap charge in a given season is the sum of a few components. His base salary for that year. A prorated slice of any signing bonus. Certain roster and workout bonuses. And any incentives the league classifies as likely to be earned, based on whether the player achieved the equivalent target the previous season.
Base salary is straightforward: paid in that year, charged in that year. The signing bonus is where the system opens up.
Signing bonus proration, and the five-year rule
A signing bonus is paid, in cash, more or less immediately. It is the money that makes a player sign. It is also, from the club's point of view, gone: once it is paid it cannot be recovered except in narrow circumstances involving misconduct.
But for cap purposes it is not charged when it is paid. It is spread evenly across the years of the contract, up to a maximum of five. This is the single most important rule in NFL roster construction.
Consider a straightforward deal. A club signs a player to four years, with a $24 million signing bonus and base salaries of $2 million, $10 million, $16 million and $18 million. The cash the player receives in year one is $26 million. The cap charge in year one is $2 million of base salary plus $6 million of prorated bonus, which is $8 million.
The club has paid twenty-six and charged eight. The other eighteen has not vanished; it has been pushed into the future, six million a year for three more years.
Read the two bars against each other and the whole strategy of NFL contract structuring is visible in one picture. Year one is the year the club buys cheap. The bill arrives later, in years three and four, by which point the cap has grown, the club has renegotiated twice, and in a great many cases the player is no longer on the roster.
The five-year ceiling matters because it caps how far the pushing can go. A seven-year contract does not get seven years of proration. The bonus still divides by five. This is why very long NFL contracts are usually four or five years of real football with paper years attached, and it leads directly to the void year, which we will come to.
Dead money, and what a release actually costs
Now the other half. A club decides it no longer wants the player. It releases him. What happens to the proration it has not yet charged?
It accelerates. Every remaining dollar of prorated signing bonus hits the cap immediately, in the year of the release. That charge is dead money: cap space consumed by a player who is not on the roster, for cash the club paid out years earlier.
Take the contract above and release the player after year two. He has been charged $6 million of proration in each of two years, so $12 million remains unamortised across years three and four. Release him in the year-three offseason and that $12 million lands on the year-three cap in a single lump. Against this the club saves the year-three base salary of $16 million.
Net saving: $10 million. The club is better off, but it has spent $12 million of cap space on a player who will not take a snap for it.
This is what "dead money" means, and it explains something that confuses people every March. When a club is reported to be carrying a huge dead money figure, that is not a debt. It is not money it still owes. It is the accounting tail of cash that was paid, and enjoyed, in previous seasons. A team with $50 million of dead money has already had the benefit. It is paying for it now.
It also explains why the most damaging contract in football is not the expensive one. It is the expensive one signed by a player who declines fast. A club can absorb a large cap charge for a good player. What it cannot easily absorb is a large acceleration for a player it needed to move on from two years ago.
The June 1 designation
There is a valve. A release processed after 1 June splits the acceleration: only the current year's scheduled proration is charged this year, and everything beyond it moves to next year.
Because clubs need to make roster decisions in March, not June, the agreement lets each club designate up to two releases per year as post-June 1 for cap purposes, while actually releasing the player in March. The player is free to sign elsewhere immediately. The cap relief, however, does not arrive until 1 June, which means the club cannot spend it during the opening week of free agency, when the best players sign.
That trade is the point. The June 1 designation converts a large single-year charge into two smaller ones, at the cost of timing. Clubs use it when they are certain, and they get two of them a year.
- The club decides to move onThe player's remaining base salary comes off the books. Cash already paid as signing bonus does not come back.
- Unamortised bonus is identifiedEvery dollar of signing bonus prorated into future years is still owed to the cap. This is the number that becomes dead money.
- Route A, a standard releaseThe whole remaining balance accelerates onto this year's cap at once. Simple, immediate, and expensive in a single season.
- Route B, a post-June 1 designationOnly this year's scheduled slice is charged now. The rest moves to next year's cap. Two of these per club per season.
- The relief landsOn a standard release, straight away. On a June 1 designation, not until 1 June, which is after the free agents worth signing have signed.
The same release, taken down two different routes. Which route a club picks is a timing decision, not a money decision: the total charged is identical either way.
Restructures, and the void year
The other lever is a restructure, and it is the one that makes "cap hell" so consistently overstated.
A restructure, in its ordinary form, is not a renegotiation. Nobody takes a pay cut. The club converts a chunk of a player's base salary into a signing bonus and pays it immediately. The player receives the same money in the same year, slightly sooner. But because the money is now a bonus, it prorates.
Convert $15 million of base salary in a contract with three years remaining and $10 million of the charge leaves the current year for the two years after it. The club has created $10 million of space in an afternoon, with a phone call and a signature, and without losing a player.
The catch is that it has also lengthened the tail. Every restructure increases future dead money, because it increases the unamortised bonus that will accelerate if the player is ever released. Restructure a contract three years running and the club has quietly converted a flexible arrangement into one it can no longer get out of.
The void year takes the same idea further. The club adds contract years that both sides know will never be played, purely so that the proration divides by a bigger number. A three-year deal with two void years attached prorates the bonus across five. When the void years arrive, the contract ends and every remaining dollar accelerates.
Void years are not cheating. They are explicitly permitted, the acceleration is certain, and the league office signs off on the paperwork. What they do is exchange present flexibility for a fixed, dated, unavoidable future charge. A front office using them is making a bet: that the cap will be larger when the bill arrives, and that the intervening seasons will be worth it.
Sometimes that bet is exactly right. A club with a quarterback on a rookie contract and a defence one player short is correct to push money forward, because its window is now and the cost lands after the window closes. Sometimes it is a club with no plan, buying one more mediocre season at the price of three.
- Signing bonus, paid at once24m
- Guaranteed base salary20m
- Non-guaranteed years46m
Illustrative shape rather than a specific deal. The guaranteed portion is what the player can bank; the rest is a club option exercised one year at a time.
Show the numbers
| Item | Value |
|---|---|
| Signing bonus, paid at once | 24m |
| Guaranteed base salary | 20m |
| Non-guaranteed years | 46m |
Carryover, the spending floor, and the rule of 51
Three smaller mechanisms round out the picture, and each corrects a misconception.
Unused cap space carries over. A club that finishes a year under the cap can roll the remainder into the next year, provided it files the paperwork on time. This is why the number quoted in March is often the adjusted cap: the league figure plus that club's carryover, minus whatever it is already carrying. Clubs planning a big offseason will deliberately underspend the year before in order to bank space.
There is a floor as well as a ceiling. A club cannot simply pocket the difference. The agreement requires each club to spend a minimum proportion of the cap in cash, currently ninety per cent, measured in aggregate across multi-year periods rather than season by season. A club that finishes a period below the floor pays the shortfall directly to the players who were on its roster during it. The floor is why perennial non-spenders eventually have to sign somebody.
Only the top 51 contracts count in the offseason. Rosters run to ninety players between March and the end of the preseason, but only the 51 largest cap charges are counted until the regular season begins. This is why signing a minimum-salary depth player in May often costs a club almost nothing: his charge simply displaces the 51st, and the club is billed the difference. When rosters cut to 53, everything counts again.
The rookie wage scale, and why the draft is the real currency
Before 2011, drafted players negotiated. The results were absurd, and the reason the system changed is that the absurdity all pointed one way: unproven players at the top of the first round were among the highest paid in the league before they had played a down.
The current agreement replaced negotiation with a scale. Each draft slot carries a substantially predetermined value, contracts run four years, and first-round picks carry a club option for a fifth. That option, since the most recent agreement, is fully guaranteed when exercised and is priced in tiers according to how much the player has actually played and whether he has been selected for the Pro Bowl.
The consequence is the single largest structural force in the modern NFL. A quarterback taken in the first round is under contract for four years, with a fifth available, at a cost that is a small fraction of what an established starter commands on the open market. A club with a good quarterback on a rookie deal has an enormous, temporary, and precisely dated advantage, because it can afford to pay market rate at every other position at once.
Everything about how modern front offices behave, the aggressive win-now trades, the void years, the willingness to carry dead money, follows from trying to convert that window into a trophy before it closes. It is also why the draft is treated as the real currency of the sport rather than free agency: a pick is not a player, it is a discount, and the discount is what buys everything else.
The parts of a cap charge that nobody mentions
Base salary and prorated bonus account for most of a cap number, but not all of it, and the remainder is where clubs do their quieter work.
Roster bonuses are paid for being on the roster on a particular date, usually in March. They charge in full in the year they are earned, and they exist to force a decision: the club must choose to keep the player before free agency opens, rather than waiting to see what else is available. A roster bonus is a deadline dressed as money.
Workout bonuses are paid for attending a set proportion of the offseason programme. They are small, and they are one of the few tools a club has to make voluntary attendance less voluntary.
Option bonuses behave like signing bonuses. They are paid when the club exercises an option built into the contract, and they prorate across the remaining years, subject to the same five-year ceiling. Contracts written with option bonuses are contracts written with a restructure already loaded.
Incentives are split into two categories, and the split is the interesting part. An incentive is likely to be earned if the player met the equivalent target in the previous season, and not likely to be earned if he did not. Likely incentives charge against this year's cap whether or not they are actually earned. Unlikely ones do not charge at all until they are, at which point the charge lands on the following year's cap.
This produces a genuine loophole and everybody uses it. A player coming off an injured season has a low baseline, so targets he will plainly hit are classified as unlikely, and the club gets the cap relief now and pays the bill next year. Agents structure around it deliberately. The league office reviews the classifications, but the rule is mechanical rather than discretionary, and a mechanical rule with a known input is a rule that can be aimed.
The 90, the 53, the 48 and the 16
Roster mechanics interact with the cap in ways that catch people out.
Between March and the end of the preseason a club may carry ninety players. From the start of the regular season it carries fifty-three on the active roster, of whom a smaller number are actually eligible to play in any given game, and a practice squad alongside it.
The practice squad matters more than its status suggests. Those players are paid, they count against the cap, and they can be signed away by any other club to its active roster at almost any time, with only a short window in which the original club may match. A practice squad is not a reserve list. It is a shop window with a queue outside it.
Injured reserve is the other half. A player placed on it is out for a defined minimum period, his cap charge stays exactly where it is, and a limited number may be returned to the active roster during the season. Injury does not create cap space. This surprises people every autumn: the club still pays, the cap still charges, and the replacement's salary is added on top.
The tags
Two more instruments deserve a mention, because they come up every February.
The franchise tag is a one-year contract a club may apply to one pending free agent. Its value is the greater of two calculations: an average built from the highest salaries at that player's position, or a fixed uplift on his previous salary. It is fully guaranteed once signed. In its exclusive form no other club may negotiate with the player at all; in the non-exclusive form another club may sign him but must surrender two first-round picks, which in practice almost never happens.
The transition tag works the same way at a lower price, calculated from a wider group of salaries, and grants only a right to match another club's offer rather than compensation.
The tag is often described as a club being unable to reach agreement. More often it is the club choosing certainty: a known one-year price, at a position where a long-term deal would carry guarantees it does not want to give. The mechanics of both instruments, and the reasons a player will sometimes refuse to sign one, are worth their own explanation.
Cash over cap, and what an owner is actually spending
There is a second set of books, and it is the one that tells you about the owner rather than the general manager.
Cap accounting spreads money across years. Cash accounting does not. A club that pays three large signing bonuses in one March has a modest cap charge and an enormous cash outlay, and the gap between those two numbers is called cash over cap.
The distinction matters because the cap is equal across the league and cash is not. Every club faces the same ceiling, but not every owner is willing to write the same cheque. A club that consistently spends more cash than cap is front-loading: paying now to charge later, which requires an owner with the appetite and the balance sheet to fund it. A club that consistently spends less is doing the reverse, and will eventually meet the spending floor coming the other way.
This is the closest thing the NFL has to a rich-club advantage, and it is real but bounded. It cannot buy a better roster this season. What it can buy is flexibility: the freedom to structure contracts in the way that suits the football decision rather than the way that suits the cash flow. Over a decade that compounds, which is why the same handful of clubs always seem to have room and the same handful never do.
If you want to see how differently other sports answer the same question, the comparison across the major leagues is the short version: hard cap, soft cap, tax, or nothing at all, with wildly different consequences for competitive balance.
Why cap hell is a myth
Put the pieces together and the February panic dissolves.
A club reported to be $40 million over the cap has, in nearly every case, a straightforward path back. It can restructure two or three large contracts and create most of that space in a week without losing anybody. It can release players whose base salaries exceed their remaining proration, which produces net savings. It can use its two June 1 designations on the worst two contracts on the books. It can decline options. Its adjusted cap may already be higher than the number being quoted, because carryover has not been included.
What the club cannot do is make the money disappear. Every one of those moves pushes a charge into a future year. The clubs that genuinely suffer are not the ones that go over; they are the ones that went over three years running, restructured everything, attached void years to the rest, and arrive at a season where a quarter of the cap is committed to players who have retired or moved on.
That is a real condition and it is worth writing about. But it is a slow, self-inflicted, entirely visible condition, and it looks nothing like the annual March headline. The headline describes an accounting position on a particular Tuesday. The real thing takes half a decade to build and about the same to escape.
The useful question about any club's cap position is therefore never "how far over is it". It is: how much of next year is already committed to players who will not be here, and how many more times can this front office push the bill forward before there is nowhere left to push it.
Reading a club's cap position properly
If you want to judge a team's finances rather than react to a headline, four figures tell you nearly everything.
The first is committed cap for next season, not this one. This season is a solved problem by the time it starts; every club gets compliant. Next season is where the choices show.
The second is dead money already scheduled. Void years and past releases produce charges that are fixed and dated. A club with a large scheduled dead charge two years out has already spent that year.
The third is how many contracts have been restructured more than once. Each restructure is a signal that the club needed space and had no better source of it. A pattern of them is the clearest early sign of a front office out of ideas.
The fourth is the quarterback's cap number as a share of the cap, and how long it stays there. Nothing else on a roster moves the arithmetic as far.
Those four numbers are public, they are stable, and they will tell you more about a club's next three seasons than any amount of March commentary about who is in trouble.
None of them is difficult to find. All of them are ignored every February in favour of a single negative figure that will be resolved by the end of the month, because the single negative figure fits in a headline and the four that matter do not.
More of this sort of thing, on this sport and others, is collected in the American football archive.
Common questions
Is the NFL salary cap a hard cap?
Yes. Unlike Major League Baseball, which has a tax and no ceiling, or the NBA, which has a soft cap with a long list of exceptions, the NFL cap is a genuine limit. A club must be under it at the start of the league year and must stay under it, and there is no fee a wealthy owner can pay to exceed it.
What is dead money in the NFL?
Dead money is salary cap charge for a player who is no longer on the roster. It is almost always unamortised signing bonus: money the club has already paid out, which it was spreading across future cap years, and which accelerates onto the books the moment the player leaves.
Do NFL teams have to spend all of their cap space?
Not in a single year, but they cannot hoard it indefinitely. The agreement sets a cash spending floor across multi-year periods, currently 90 per cent of the cap in aggregate, and a club that finishes below it pays the shortfall directly to the players who were on its roster.
What is a void year?
A void year is a contract year both sides know will never be played. It exists so that a signing bonus can be spread across more seasons, lowering the charge in each. When the year voids, every remaining dollar of proration accelerates onto the current cap.
Can a team go over the cap?
Only in narrow, temporary circumstances during the season, and never at the start of the league year. A club that cannot get compliant faces forfeited draft picks and fines, which is why the deadline is met every single year without exception.
Filed under American Football·nfl · salary cap · contracts · roster building · cba