Getting partner compensation "right" is an issue I have, and will continue to, recur to.  Why?  Essentially because I believe that the two extreme models (the strict lockstep and the strict "eat what you kill") each works in only the rarest of firm environments, and that, simple and straightforward as they may be to implement and understand (with the concomitant benefit that partners know what they're signing on for), they will create a "disequilibrium condition" in the vast run of firms.

Let's pause a moment to review the bidding:

• In favor of lockstep models are that they:  Promote collegiality; encourage sharing of clients and assigning work to the lawyer or practice group most suitable regardless of origin of the "relationship;" and, by eliminating subjectivity, eliminate second-guessing and much of the opportunity for perceived grievances about pay.
• In favor of eat-what-you-kill is that they:  Strikingly encourage entrepreneurship and business development; have an appealing meritocratic aspect; strictly penalize slackers and "free-riders;" and, arguably, spur overall growth of the firm faster than any alternative model.

The demerits of lockstep and of eat-what-you-kill are essentially the inverse of the other's virtues:  For example, if lockstep promotes collegiality, eat-what-you-kill promotes Lone Rangers whose professional focus is (understandably!) on their book of business—often a portable asset.  Lockstep will work in firms with exceptionally strong traditions (think Davis-Polk, with the remarkable, and so far as I know unique, record of never having suffered a partner defection to another firm).  Likewise, eat-what-you-kill probably works in young and hungry plaintiffs' law firms (a market segment you have never heard mentioned here before and, I'll lay odds, will never hear mentioned again).

So then, what is the alternative model for the vast run of firms?

It is, I suggest, a benevolent and enlightened form of somewhat-secret paternalism.  What?!  Am I dissing the new governance god of "transparency?!"  Indeed I am.

The model I'm suggesting has these characteristics:

• the managing committee, or, if that's too unwieldy, a subset thereof, sets each individual partner's compensation annually, and does so on a subjective non-numeric basis ;
• this is based on a full review of the entire spectrum of each partner's contribution to the firm during the previous year (aided by each partner's own "annual report" on herself);
• critical areas of contribution include, of course, billing and business development, but also associate development, pro bono or community activities, participation in practice group or firm management, publications and presentations;
• no partner knows the exact compensation of any other partner (save, of course, those within the managing committee), but the partnership as a whole is presented with a thorough overview and discussion of the range of compensation, the factors included and their relative importance given the past year's business climate and the firm's strategic objectives.

It is a truism that, in the long run, you get the behavior you encourage.  To encourage both business-building (an outward-looking metric) and cross-firm collegiality (an inward-looking virtue), one has no choice but to provide incentives for both. 

Got a better idea?