I don’t normally like to just share  other people’s content without expanding upon what they’ve said.  But this post over at AdamSmithEsq. is really fantastic and I don’t think there’s much I could add to make the point any better.

I suggest that afterwards, you take a few minutes to revist my main website and reconsider whether and to what degree you have ever been a victim of the Doctrine of Sacrifice in your own career or life.

The following is from Bruce over at  Adam Smith Esq, a blog I highly recommend:

I want to discuss a theory floated by two economists posing an alternative explanation for large firms’ “up or out” partnership structure, which was covered   in The Wall Street Journal last week.

As you know, the conventional explanation for what is essentially a binary personnel policy—associate or partner—is usually referred to as the “tournament” model, positing that associates compete to win the “tournament” of election to partnership, and that both the carrot of tremendous rewards should they succeed, and the stick of unconditionally being forced to leave if they do not, provide ample incentive for them to work, shall we say, with extreme diligence for 7 to 10 years.

In lieu of that explanation, the professors theorize   that the real explanation has to do with “property rights,” and specifically with the one key asset the firm has, its relationships with its clients.  After making the commonplace observation that law firms have no material physical assets, they hypothesize:  “The one-sentence summary of the model is when knowledge assets are very valuable, the structure of the organization is going to adapt to try and take advantage.”  More specifically, the risk to any firm is of course that people can walk out of the firm with a critical asset—a client relationship—and one way to control or minimize that risk is to put very tight limits on membership in the circle of people who truly have high-quality contacts with clients and who therefore pose a threat of taking them away.  Thus, a small partnership cohort vis-a-vis the firm as a whole.

Add to this the practice of keeping associates effectively at arm’s length from clients and, our professors theorize, the firm has as solid a handle on clients as feasible.  As one reader put it to me in an email:  “Once associates get too experienced, keeping them around as anything less than partner is too dangerous. It seemed pretty persuasive to me, at first glance.”

The problem is that 80% of the AmLaw 200 do precisely that, as document in Prof. Bill Henderson’s paper   on Single-Tier vs. Two-Tier Partnerships in the AmLaw 200 (80% of AmLaw 200 firms have a non-equity “partner” tier).  How, then, do the professors deal with this?

Essentially, through flatly denying reality.  With straight faces, they say:

You’d think the second prize, instead of getting fired, would be a somewhat smaller salary than being partner. And firms don’t do that, which may seem surprising. To us, it’s not surprising at all. The worst thing you could, in our conception, would be to keep a bunch of people around who are competent but underpaid because they’d be the ones who walk off with all the clients.

And when called on this contradiction by the WSJ interviewer, who asks “How do these people [non-equity partners] fit in?,” they demonstrate a basic misunderstanding of the role non-equity partners play, describing them as “people that are very good at some aspect of the job,” and nothing more.

Now, is this a fatal defect in their theory?

My take is that it doesn’t entirely knock it off the table, but it certainly is not going to dethrone the “incentives/tournament” theory any time soon in my mind.  Yes, firms obviously know that their only assets are client relationships, and as the shockingly rapid implosion of some firms over the past decade has shown, those assets are highly portable and the firm’s hold on those assets can be fragile.  But I think firms respond to that more by how they manage their own law firm partnership structure than by the far blunter instrument of the up-or-out structure.  Which, we know, only holds true for 20% of the AmLaw 200 today.