Free agency in the National Football League began this month and a number of stars—Ndamukong Suh, DeMarco Murray, and Darrelle Revis—have found new employers. NFL fans everywhere love looking at each move their favorite teams make and debating whether or not these moves are likely to improve their own future happiness.

Although people might disagree about the merits of each transaction, there is a general sense that each team is trying to win. After all, don't teams have a clear financial incentive to win more games? Don't more wins lead to more fans in the stands and more viewers on television? And doesn't all that make the team owners more money?

As it turns out, the economics of the NFL don't quite work this way. The NFL equally shares its nearly $5 billion of national television revenue among all its teams. It also shares a substantial portion of its ticket and merchandise revenue, but not revenue from suites, sponsorships, or naming rights. All of this means that the link between a team's record and the revenue it brings in is quite weak.

In a recent paper published in the International Journal of Sport Finance, Michael Leeds, Peter von Allmen, and I look at the statistical link between a team's wins and its total revenue in the NFL, National Basketball Association, and Major League Baseball. With respect to baseball we found that a 10 percent increase in regular season wins for an average team would lead to a 2.7 percent increase in revenue. The same result was uncovered for the NBA. In both of these leagues the national television revenue is shared, but other revenue streams, such as local media, gate revenue, and sponsorship revenue are—relative to what we see in the NFL—not shared as much.

In the NFL, by contrast, a 10 percent increase in regular season wins for an average team only leads to a 0.14 percent increase in revenue. Because the NFL has embraced much more sharing, the financial incentive to win is muted. The impact of wins in the NFL is only a small fraction of what you see in the other two major North American sports.

These results can be better illustrated by thinking about a player's economic contribution to his team. In 1974, sports economist Gerald Scully wished to understand how the inability of players to bargain with other teams (baseball did not have free agency in 1974) impacted the wages paid to the players. To address this issue, Scully argued that a baseball player's economic contribution could be measured by noting how a player's actions impacted wins and what those wins were worth in terms of revenue.  

For example, Reggie Jackson was the American League MVP in 1973. That season, Jackson was worth about eight wins for the Oakland Athletics. An examination of the link between wins and revenue reveals that a win, in 2015 dollars, was worth about $900,000. So Jackson was worth between $7 and $9 million to the A's in 1973 (again, in 2015 dollars). Because baseball players in 1973 had no free-agent rights, Jackson was paid quite a bit less than what he was worth. His salary was only $70,000, or less than $400,000 in 2015 dollars; his lack of bargaining power meant he was paid dramatically less than his value to the team.

Now let's apply the Scully approach to football players today. An examination of the link between revenue and wins in the NFL reveals that each additional victory is worth about $412,000 to an NFL team.  That's right: A win in the NFL is actually worth less, in real terms, than a win in baseball more than 40 years ago.

To put the NFL number in perspective, consider two teams. The first wins twelve games; the second wins eight, making it average. These results indicate that the 12-win team—relative to an average team—will earn less than $2 million more in revenue. An average NFL team, though, earns more than $250 million in revenue. So more wins really doesn't dramatically change a team's financial picture.

And that means that a player like Aaron Rodgers (following the Scully method) appears to be fantastically overpaid. According to Advanced Football Analytics, Aaron Rodgers (the 2014 NFL MVP) had a 5.87 "wins probability added" last season. Given the small amount of revenue generated by one win, this means that (again, according to the Scully method) Rodgers' economic value was less than $2.5 million last year—a sum that is substantially less than $14.5 million Rodgers received from the Green Bay Packers last year.

So does this mean Rodgers is in fact overpaid? Not exactly. Since an NFL team's wins do not have much impact on a team's revenue, no matter what Rodgers does on the field, he doesn't have much impact on the revenue of the Green Bay Packers. And this means that economists are not really able to ascertain the economic value, via the Scully method, of Rodgers or anyone else playing in a league in which the majority of revenues are shared.

If teams have little incentive to win, why do teams spend much money on players at all?  In the past, this was more of an option. But today the NFL not only has a cap on payrolls, it also has rules in place that require teams to spend a certain amount on players. So teams have some ability to lower spending, but there is a limit to how little they can choose to spend on payroll.

So why don't all teams simply spend as little as the league allows? One suspects that teams are not strictly motivated by financial incentives. People in the NFL don't try to win just to make money. They also are motivated to win games because they like winning.

But although winning can make people in the NFL happier, the big economic story is that winning doesn't make owners in the NFL much richer. In sum, teams simply do not have much of a financial incentive to win. No matter what happens, next season there will be equal numbers of wins and losses across the league, because for every team that wins, another one loses. But not so for the owners: They, unlike their teams, will all get to be winners.