Charlie Munger created a crazy incentive system
And it helped make Munger, Tolles & Olson the best law firm in the country
Commoncog had an interesting post recently about the incentives at Charlie Munger’s law firm.
Each partner fills in the amount of money he or she thinks every participant should make, with no rules other than that the numbers must add up to the net income for that year. “There are no points, no shares, no extra credit for seniority. When the partners vote, they can take seniority into consideration, but they also consider a person’s ability to find business and represent clients successfully. ‘“We vote on disbursement, then everybody gets to see how everyone else voted,” explained Olson.
Munger, Toles & Olson as its been an incredibly successful law firm. From the post:
Munger, Tolles & Olson has topped The American Lawyer A-list a record nine times, and maintains the highest average A-list ranking in the history of the award.
Barring significant counter evidence, I think we have good reason to believe that the incentive system works well for them, including
Munger Tolles enormous success.
The long length of time it’s been in use. After nearly 60 years if it was not working well it would have been modified heavily or discarded.
The social difficulty of the system, that forces partners to do something they normally don’t like to do, demonstrating that most think its worth the emotional effort.
My experience is that people hate publicly criticizing co-workers. Even managers usually struggle to give well deserved reprimands. We have an inherent need to be liked, and criticizing others opens the door for others to dislike and even criticize us. The fact that Munger, Tolles & Olson continue what is such a humbling and difficult process for all these years is strong evidence they find it extremely valuable.
But what most caught my eye was a comment musing about whether it was applicable to a technology company. More specifically I wondered if it would work in an engineering organization. In both cases, I’ve come to a tentative opinion that it wouldn’t.
The reason I think an organization wide public grading system would fail is the hierarchal organizational structure nearly all tech companies are based on. I think our innate bias against publicly criticizing others becomes extreme if asked to publicly criticize higher ranking employees. I have seen many employees with legitimate complaints about a manager or their decisions freely share their gripes with peers in private, but in a public forum when when asked for any concerns they demur and say they don’t have any. No one wants to paint a bulls-eye on their back for a more powerful manager to target. So any open public grading system seems likely to end up with managers getting graded far too leniently.
Aside: This is a general problem in every business. As a leader you should want to cultivate a culture where direct and honest feedback is always expected regardless of rank. But its a very hard thing to do. Thats one reason management techniques such as “management by walking around” have so much power, if you are a high ranking executive in a large organization much of the critical information you get has been filtered through layers of managers. It’s not just the telephone game, it’s the telephone game where every operator has a strong self-survival bias to ensure their boss gets a rosy picture of their contributions. But when a leader uses MBWA to create opportunities for private conversations with people many levels beneath them in the hierarchy (colloquially known as” the people who do the actual work”), often they are able to hear unvarnished truth.
I thought about overcoming the hierarchical issue by applying this system within single ranks, such as all non-managerial engineers, or even more specifically such as all senior engineers. But the problem is, how much does one engineer really know about what other engineers in their organization are doing, and how well they do it? In a large organization, very little other than those whom you interact with directly on specific projects. In that small group you likely can identify superstars and laggards, but again bias rears its ugly head. It’s easy to praise someone who is always helpful and does great work on time. But it’s much harder to publicly criticize one who isn’t helpful, and is always slow or late. Emotionally it’s much easier to tell your manager, and hope they can work it out with the “problem”s manager, hoping to spare you a confrontation.
So right now the only way I can see to apply it within a tech company is to apply it within teams. They would have to be small enough so that everyone works closely enough together to have reasonable opinions about each other’s contributions. And they would need to be peers, so no managers. So in most teams would be left with some pretty small grading groups.
To summarize, I really like this system. I think it’s clearly effective for Munger, Tolles & Olson and has likely strongly helped make them the huge success they are. Truth isn’t just important for good managerial decisions, it’s critical. So I’m always interested in tools that make (or keep) organizations honest. But so far I don’t think a public open grading system would work well in the far more common hierarchical organizations, unless it is narrowed down to use in groups of equals with enough information on each other’s performance to make it both honest and fair.