Last week’s episode of The Good Wife (““Trust Issues”) was interesting for two reasons: it used a “ripped from the headlines” legal case that I discuss in my book on bargaining and the legal argument they use is essentially a trivial application of pre-play cheap talk in a repeated prisoner’s dilemma.
The $9 Billion Google/Apple Anti-Trust Lawsuit
First, the background of the real life version of the case. In the early 2000s, Google and Apple (along with Adobe and Intel) allegedly had a “no poaching” gentleman’s agreement. That is, each company in the group pledged to not attempt to hire employees at any of the other companies. The employees eventually figured out what was going on, filed a $9 billion lawsuit, and settled in April 2014 for an undisclosed amount.
Why is the practice illegal? It goes without saying that quashing competition among firms hurts the employees’ bargaining power, and the law is there to protect those employees. But what is not so clear is just how attractive a no poaching agreement is to the firms. In fact, when companies play by the rules, just about all of the potential for profit goes into the employees’ hands.
To see why, imagine that Google and Apple both wanted to hire Karen. Karen has impressive computer programming skills. And because Google and Apple value computing skills at a roughly equal rate, suppose that the most Google would be willing to pay her is $200,000 while Apple’s maximum is $195,000. Put differently, $200,000 and $195,000 represent the break even points for the respective companies. Put differently again, Karen will bring in $200,000 in profits to Google and $195,000 to Apple, so hiring her for any more than that will result in a net loss.
How will that profit ultimately be divided between Karen and her employer? You might think that Google should be the one hiring her. And you are right—she is worth $5000 more to Google than Apple. You might also think that Google will profit handsomely from her employment. However, as I discuss at length in the book, the logic of bargaining shows this to be untrue. If Google offers Karen any less than $195,000, she can always secure a job from Apple; this is because Apple values her at that amount, and so Apple would be willing to slightly outbid Google to hire her. Thus, the outbidding process ultimately ensures that Karen receives at least $195,000. She is the real winner. Although Google might still profit from her employment, its net gain will not exceed $5000 ($200,000 – $195,000).
So the firms have great incentive to collude, drive down wages, and secure more of the profits for themselves. What does that sort of collusion look like?
Well, we might think of it as a repeated prisoner’s dilemma. In this type of interaction, in any given year, each of us would maximize profits by trying to poach the rival firm’s employees regardless of what the other firm chooses to do. (If you don’t poach, then I make out like a bandit. If you do poach, I’m still better off poaching and not losing all the employees.) However, because each of us is poaching and driving up employee wages, both us are ultimately worse off than if we could enforce an agreement that required us to cooperate with each other and not poach.
Of course, anti-trust laws prevent us from explicitly contracting such an agreement in a legally enforceable manner. However, an informal and internally enforceable agreement is possible. Suppose we both start off by cooperating with each other by not poaching. Then, in each subsequent year, if both of us have consistently cooperated before, we continue cooperating. Otherwise, we revert to poaching.
Would anyone like to break the agreement? No. Although I could gain a temporary advantage against you by poaching your employees today, the higher wages over the long-term with mutual poaching are going to vastly outstrip that short-term benefit.
This is exactly the type of agreement Google and Apple struck. In fact, when a Google recruiter attempted to hire some Apple employees, Steve Jobs shot the following email to Google bigwigs: “If you hire a single one of these people, that means war.”
Alicia Florrick’s Defense
The episode of The Good Wife featured fictionalized versions of Google and Apple involved in the same affair. Like reality, employees caught on and sued.
The plaintiff’s lawyers thought they had the case in the bag. Indeed, they had turned one of the owners of a trust company against the defense. He went on record that the defense had negotiated the terms of the no poaching policy explicitly and was very happy to agree to the deal.
Alicia Florrick (the defense attorney and titular Good Wife) had a great defense: any discussion of such an agreement is not an unambiguous signal of plans to break the law. These repeated prisoner’s dilemmas have an interesting property in that regardless of whether you plan to cooperate with the other company or screw them over at the first possible moment, you always want to convince the other side that you will cooperate. If you plan to cooperate, then you want to tell the other side to cooperate as well so you can sustain that cooperation in the long term. If you want to follow the law and poach freely instead, you still want to convince the other side that you are going to cooperate so that they cooperate as well, allowing you to screw them over in the process.
So Florrick points out that this type of pre-play communication is meaningless. Regardless of the ultimate intend, the defendant would say the exact same thing. The testimony therefore proves nothing. The plaintiff promptly settled.
All told, I really appreciate two things about the episode: its sophisticated understanding of a potentially very complicated strategic situation and the how punny the “Trust Issues” title is.