It is fast becoming an imperative for elite firms to widen the range of their partner compensation. Too narrow a range allows competitors with wider ranges to lure away the most commercially successful partners. We saw this in London when the U.S. firms arrived and undid the elite London firms’ lockstep models. We are seeing this increasingly in New York where firms like Kirkland & Ellis, who reportedly moved recently to a ratio of the compensation of their highest- to lowest-compensated partners of 9-1, pose a renewed threat to old-line firms with narrow, 3-1-type ratios.
There are a number of reasons to believe that a compensation range of about 9-1 is consistent with the range in economic contribution of individual partners’ practices and is thus the range a firm’s compensation must reflect to avoid having its partners be cherry picked by others. One is that 9-1 is consistent with what I’ve seen at elite professional services firms as a consultant—the rule of thumb I had developed was that newly-promoted (and hence lowest-comped) partners earned about one-third of firm average while the most commercially productive partners receive three times firm average—from one-third to three times firm average is a 9-1 ratio of top to bottom. I should note that here, and in what follows, the ratio I refer to is that between the average compensation of the top and bottom deciles of partners by compensation, and not of the highest- and lowest-compensated individual partners.
The fundamental driver of wide compensation bans is that, even among teams of elite global professionals, there is variation in the intrinsic value of the roles they play, how well they play them, and where individuals are in their careers in terms of their bargaining power. This variation in value is reflected in the compensation of the professionals. There are few professional markets that are unconstrained and transparent. One is that of British premier league soccer players (U.S. sports markets are distorted by league minima, restricted free agency, etc.). The table below shows the compensation ratios of the teams with the highest total wages. What you see is a broad range, but none has a ratio below 12-1.
A third reason supporting a wide compensation ratio is that some pretty savvy law firms are operating at this type of range. Kirkland & Ellis, for example, has been very successful in expanding from its Chicago roots to New York, national, and international prominence. They also exhibit low lateral partner attrition. It’s hard to imagine such success if the compensation model were significantly out-of-sync with economic contribution.
Winners and losers as the range expands
Going from a narrow to a wider compensation range is challenging at many levels. Not the least is that it creates winners (those who earn more) and losers (those who earn less). Chart 1 shows the partner compensation distributions at two model firms. Note that the distribution of partner compensation is not modeled as a symmetrical bell curve but rather as a skewed (technically a log-normal) distribution. This is consistent with the many partner compensation distributions I have seen. The curves on Chart 1 show the distributions for firms with the same average partner compensation of $2 million, but with two different ratios of high-to-low compensation—3-1 and a 9-1 ratios. What you see is that as you go from a 3-1 to the 9-1 ratio, and the compensation of partners at the high end goes up (A), the compensation of the bulk of partners has to go down (B) to offset the increases.
Chart 2 shows the same data but on a cumulative compensation basis, i.e., the y-axis indicates the percent of partners who receive less than a particular compensation level on the x-axis. The chart shows that in going from a 3-1 ratio to a 9-1 ratio, one-third of partners would see their compensation increase while two-thirds would see their compensation decline. Hence, the challenge for elite law firm leaders: while widening compensation distribution may be rational and necessary to protect a firm’s most commercially-successful partners and hence the firm’s financial strength, it’s not going to be popular.
The keys to selling internally a new compensation system are two-fold: a new design that is comprehensible and rational, and a redesign process that is transparent.
A rational approach to compensation redesign
Most firms’ compensation systems today involve assigning to partners designated portions of the aggregate profit pool (using share allocations or some such mechanism). This makes widening the range problematic—it could appear arbitrary, confuse partners as to what is being rewarded, and open the floodgates on hurt feelings around hard-to-explain relative movements.
Instead, I recommend moving to a system that breaks the aggregate profit pool into buckets to reward partners for different contributions and then allocates these buckets to partners in a manner that is consistent with what is being rewarded and can be readily articulated. For example, as shown in the schematic in Chart 3, the aggregate pool could be broken in to four buckets to reward partners for:
3. Profitable execution, and
4. Getting along.
The ownership piece rewards partners as owners rather than for year-by-year performance. Allocating about 20 percent of the total profit pool on this basis is probably about right. The allocation of this 20 percent to individual partners should be related to their long-term contribution in building the firm, e.g., based on some kind of lockstep or, say, each partner’s individual cumulative years as a partner as a percent of the total across all partners.
The origination piece rewards partners for bringing in work irrespective of who executes it. Somewhere between 25 and 30 percent of the profit pool could be allocated on this basis. The allocation to individual partners is something partners should negotiate matter-by-matter, but do so within a transparent framework that prescribes norms for particular situations, e.g., for a long-term client (where some of the credit goes to the long-ago originator of the relationship); a new client; a client brought in by a lateral partner, etc.
The volume of work originated should not be measured in gross revenues. One should look at least to net revenues (i.e., revenues after discounts and realization losses) or, better, cash collected. Even better is to use contribution margin, i.e., cash (or net revenues) less standard costs for the non-(equity)-partner timekeepers. The standard costs would be based on the approximate hourly costs of the nonpartner time, i.e., billing rate divided by the firm’s ‘multiplier’ or ratio of billing rate per-hour to salary converted to an hourly equivalent. Using contribution margin gets past the variation in profitability caused by the different levels of resources that different matter types consume and thus focuses instead on the amount the matters actually contribute (hence ‘contribution’) to coverage of the firm’s fixed non-timekeeper costs.
The profitable execution piece needs to be looked at in a way that puts on a comparable basis high-realization, low-leverage, corporate work and low-realization, high-leverage, litigation work. The way to do this is to look at contribution margin per partner hour. Indeed, the profitability of an individual partner’s execution hours can be determined by adding up the partner’s hours on each client/matter times the margin per partner hour (MPH) of those clients/matters. Again, allocating 25 to 30 percent of the total profit pool on this basis seems about right to me.
The final ‘getting along’ piece rewards partners for playing well together, i.e., collaboration, developing firm relationships constructively, building the franchise, supporting each other’s growth, etc. Again, about 20 to 30 percent of the total profit pool could be allocated on this basis. The allocation could be based on how a partner is rated by her peers for such behavior; however, this approach tends to beget gaming. An approach I have seen be more effective is to let practice group and office leaders do the allocation based on a simple grading system (with varying compensation by grade) and subject to a forced distribution curve (to avoid grade inflation).
The percentages by which the aggregate pool is broken into buckets will depend on a firms’ priorities at a particular point in time. In the same vein, a fifth bucket could be added for, say, marquee matters, innovation, global teaming, etc. as business priorities suggest. A curious thing about this kind of disaggregated mechanism is that it leads naturally to a wider range in compensation than tends to evolve from a simple one-step allocation of the aggregate pool.
Getting from here to there
Changing a compensation system tends to make partners anxious. The best way to do it is through a slow, open, process with multiple opportunities for partner engagement. Hence, my recommended approach is a year-long multi-phase process with intensive partner dialog and engagement at each stage. The transparency helps put partners at ease, and the protracted time frame allows partners get used to the idea of a change (often a more difficult step than accepting the actual change). The process also allows firm leaders to gauge the mood of the group with respect to the degree of change, pick up some better ideas to include in the redesign, and enhance their positions as thoughtful, responsive, leaders.
The process is structured with a preparatory Phase Zero and then four phases that would proceed at the pace of a calendar quarter per phase. My bias is to have firm leaders pass control of the process to a steering committee. It’s important symbolically: it conveys the notion that the new system is being developed by the partners for the partners (and not being imposed top down). The committee should comprise members across offices, practices, and tenures—I should acknowledge a personal bias though: weighting the mix to the high-potential younger partners adds a certain dynamism and is a subtle reminder to older partners of their stewardship obligation. The steering committee should be 1.5 to 2 times the number of people you think you need to carry the work load. This helps both with representation on the committee and in accommodating the reality that a number of committee members will be too swamped with the day-to-day to engage effectively. Strong staff support is vital to the committee, and outside experts can help, but the partners must always be the outward face of the effort. This includes giving the process updates and leading the small group discussions. I strongly encourage facilitation training for those who’ll lead these discussions.
Having leaders defer to the committee doesn’t mean that leaders aren’t involved. Rather, as the committee orchestrates an all-inclusive process, leaders should focus on a parallel shadow process with the most senior and most commercially successful partners (the “lions”) to ensure they feel attended to and to thereby increase the chances that they will themselves, and encourage others to, go along with the change.
I recommend that once or twice in the process you undertake a partnership-wide survey. Its primary purpose isn’t to learn something new, although that could happen. Rather, it will be further symbolism of the effort being bottoms up and, even more importantly, creates a data set that can be used to show to the partners that they are prepared to move in the directions being proposed.
You shouldn’t expect that the process will go perfectly smoothly—if it does, it’s because people are not engaging and a reset is needed. There will be many frustrations. This is normal. Over time, things will improve and confidence will build in the shared change agenda.
Perhaps the most difficult part of addressing compensation for elite law firm leaders is that while doing so is important, it’s never urgent. It takes foresight and fortitude to address it. Sadly, the alternative is to mimic the metaphorically slowly-boiled frog.