A team with a player in his "arbitration years" has essentially three options. First, it can go through the arbitration process (using what is known as "Final Offer Arbitration"). Second, it can reach a contract to avoid arbitration, whether it is a long term deal or merely a one-year deal that splits the difference between the amount the player seeks and the team has offered. Finally, the team can non-tender the player, which is shorthand for saying that the team declines to offer the player a contract at all, making him a free agent.
"What kind of crazy team would ever do this?" you may be asking. Plenty of them, argues Buster Olney:
Baseball's financial structure appears to have reached a tipping point that can be defined simply. "The arbitration process is now outdated," said a highly ranked executive, "because the players can get more money in arbitration than they would through free agency."
So now teams are about to adjust to this reality, and this is why multiple general managers expect that dozens of young players with three, four and five years of major league experience will be cut loose rather than offered arbitration in the next 41 days. Not a handful, but dozens.
You can find a speculative list of potential non-tender candidates here. What is to be done about this coming avalanche of non-tenders?!
Let's begin with the example Olney gives from the outset of his column: J. J. Hardy. In fairness, the column was published last Sunday, and trade talks are notoriously difficult to predict. Nevertheless, we now know that Hardy was traded to the Twins for Carlos Gomez.
However, the fact that Hardy was not non-tendered does not necessarily mean Olney's front-office type (FOT) was wrong. What, then did he mean?
Certainly, it would be irrational for a team to submit to the arbitration process, which can be very unpredictable given the all-or-nothing style of arbitration employed, if they could simply rehire the player on the free agent market for less. But that set of circumstances is actually quite rare. Why?
Expected surplus from arbitration
First, we have to assume that the expected surplus value of the player after arbitration is nearly zero. Otherwise, a team would be better off trading the player to another team (thus realizing the surplus) either before or after offering arbitration (whether it was before or after, I think, would depend on the team's risk aversion).
Put another way, we assume that in a competitive free agent market, teams will bid the price up until the expected surplus value approaches zero. So for free agency to be the superior option, the player's surplus value in arbitration must be even lower than it would be in free agency, which is to say the expected surplus value will be zero or negative.
One feature of the free agency market that has been commented on previously is the "Winner's Curse." When there are many bidders to an auction, and they have varying degrees of information about a player's abilities and value, some bidders will probably have valuations that are in excess of the true value (imagine a normal distribution about the player's true value). There is a good chance that the team who wins a player's services has paid more than the true expected value of that player, thus generating a negative surplus.
But wait, it gets better. The team that overvalues a player's services the most is not likely to be the team that non-tendered him! Because the team that non-tendered the player was able to observe him on a day-to-day basis, they will likely have the best gauge of his health and skills. In economic terms, the information asymmetry between that particular team and the player is lower than the asymmetry between that player and all the other teams individually. So in turn, the team that non-tendered the player is less likely to overbid for that player's services. Aren't they?
You may have wondered why, if a team has good information about a player's abilities, they would still go through arbitration and eliminate the true expected surplus value. The reason they might is because that other team who would have paid a premium for his services on the free agent market is still out there, and is a potential trading partner. So a savvy team might still offer a player arbitration when he would offer them negative expected value, in the hopes of trading him. Don't believe me? It has happened before.
But when the time comes to trade him, will the GM necessarily do so? I have been talking almost exclusively about expected value, which is an estimate of the player's value from the starting point of the offseason. But down the road, the possible outcomes are as numerous as the grains of sand on the beach. The player could realize his latent potential.
If he did, and the GM had traded him away, the GM might look pretty stupid, don't you think?
Behavioral economists refer to what is known as the "endowment effect," which can be stated two ways. One way of stating the effect is to say people try to avoid losses more than they seek gains. (In economic jargon, they are more risk averse to losses than to gains from their perceived starting position.)
Another, I think more lucid, way of explaining it is by describing the experiment that Duke economists Dan Ariely and Ziv Carmon did to demonstrate the effect's existence. They called up Duke students who had been entered into the lottery for Final Four tickets at Cameron Indoor Stadium. They asked them, if they had won tickets, how much they would be willing to sell them for. If they had not won tickets, they asked them how much they would pay to buy tickets. The students who had won were willing to sell for an average price of about $2400. The students who had not won the tickets were only willing to buy for about $170.
The students who already had the tickets valued them much more highly than the students who didn't have them, and this caused the radical divergence in their reservation prices.
Before a team either trades away or non-tenders a player, we have to ask whether they might value the player more highly because of the endowment effect.
I have the benefit of hindsight, which tells me that the Brewers did not non-tender J. J. Hardy. Even if they did not want to pay his salary, they were able to trade him to the Twins for a cheaper player. The reason such a trade was possible is because both players have positive net expected values. Simply because Milwaukee wanted to acquire a surplus at a cheaper price (perhaps because of its preference for a lower payroll) does not mean they would be willing to get rid of the surplus altogether.
Before assuming a team will non-tender a player, we must first ask if there is any surplus available to a potential trading partner. Then we must ask if the general manager would be willing to risk looking like a fool if the player has a career year. Finally we must ask if the team's valuations are affected by the endowment effect. If all of these bars are cleared, then perhaps we can say a player will be non-tendered.
But it seems like a stretch to argue there is a coming avalanche of non-tenders.