A loyal reader who has also become a friend writes that he has just re-read The American Lawyer's lead story on the AmLaw 100 for 2006 and he has some questions (emphasis supplied):

"Maybe I'm missing something, but the analysis seems flawed to me (though their conclusions may nonetheless be right). I know you've thought about this a lot more than I, so let me ask you. It seems to me that the comparsions the article draws between RPL [revenue per lawyer] and headcount growth are not all that helpful and may even be somewhat misleading.

"Classical micro theory would say add another lawyer as long as she contributes to profits. But the article seems to confuse - or at least be unclear about - the difference between diminishing marginal profitability per lawyer, absolute profits, and profit per partner. The profit maximizing partner would, it seems to me, keep adding associates until the last associate added provided no further profit. [...]

"Of course, getting your existing lawyers - fixed costs for purposes of this analysis since there is virtually no marginal cost if they work more hours - to do more work per head is more profitable than hiring additional lawyers.  But that does not mean adding more heads is necessarily bad, especially if heads are not fungible across work or if, in fact, the heads won't work more than 1850 billable hours."

The premise of the TAL piece, in turn, is, properly, in the lead:

"The historical data suggests that The Am Law 100, as a universe, is growing too fast in size to sustain its own long-term revenue expansion. All those additional lawyers are a drag on the growth of revenue per lawyer."

Or, consider this:

"Average billable hour statistics provide convincing evidence that last year The Am Law 100 had too many lawyers. Even at the 30 most profitable firms included in Citigroup's 2005 survey, partners and associates averaged only 1,850 billable hours."

Last time I asked people how they felt about 1,850 billable hours, it wasn't viewed as shirking, but it's certainly under the 2,000 hours taken as a benchmark these days.  Finally, we have this:

"Of the 28 firms with head count declines in 2005, 16 achieved increases in RPL that beat the Am Law 100 average of 7.5 percent."

And, did you now that 40% of absenteeism occurs on Mondays and Fridays (those two days being 40% of the workweek).  Of any 28 firms selected at random, you could expect 14 to "beat the average," by definition.  Is 16 a statistically significant deviation?

I'll have more of my own thoughts at the end.

But back to my friend's critique:  First, his observation that the piece does something of a mashup among the concepts of marginal profitability, absolute profits, and profits per partner is, essentially, well-taken, although to be fair the piece dwells more on revenue issues than profitability issues: After all, the sub-head of the article is, "Firms keep adding more lawyers, but converting bodies into revenue poses a challenge. It turns out, big isn't always better."

Second, he's absolutely correct as a matter of classical microeconomics that profit maximization occurs when one more unit of a factor of production (in this case, one more associate) provides no marginal profit. 

The nuance, in law-firm-land, is that you cannot build leverage to the sky regardless of what the accounting department tells you about the tradeoff between incremental expense and incremental revenue, because associates are not stupid and they will realize at some firm-specific, to-be-determined leverage ratio (you'll know it when you've hit it), that the brass ring of partnership has vanished so far into the distance that the premise of a tradeoff has become risible.   And they're out of there.

So the calculation has to incorporate not just current year or current quarter cash expenses, but the implicit cost of increased attrition as leverage rises (which is implicit, of course, only until it occurs, when it becomes quite explicit, even if never recognized as a cash item).

Finally, he properly recognizes that "if heads are not fungible across work," adding more heads can help significantly.  With this, no one would quibble, not even the TAL piece cited, which, indeed, pays obeisance to the universally acknowledged wisdom that firms "that are able to attract the highest percentage of rate-tolerant work, whether in the U.S. or internationally, will widen the gap between themselves and the rest of the firms in the AmLaw 100."

But there's far more to the story than that, and I wish to take issue with the piece's fundamental economic and statistical analysis.  One point where we differ is that I wish to take a longer-term perspective than their comparison of 2005 with 2004 and/or 2003.

As they rightly note:

"As a group, the firms of The Am Law 100 have never been very good at predicting how big they should be. Firms have typically grown based on past demand for their services -- which means that over the last 20 years there have been some calamitous mismatches between the supply of Am Law 100 lawyers and client demand for their time."

It's not, however, that simple.  It's not that the AmLaw 100 are particularly obtuse at gauging future demand; it's simply that we're an "elevator asset" business:  You cannot open or close the faucets of supply in a heartbeat.  This will never be a "just-in-time" supply chain.  

Hiring madly in an upswing entails the classic fallacy of buying at the top, and shedding madly in a downturn risks the opposite, of alienating talent and cutting capacity so aggressively that one is ill-positined to profit from a recovery.  (Indeed, I recently sat down with the managing partner of a tech-centric firm who was lamenting precisely the cyclicality of their business and exploring out loud opportunities to diversify in order to smooth the boom and bust swings.)

So I want to proffer not a 2- or 3-year comparison, but a 10-year comparison. 

Let's look at the 2006 AmLaw 100 vs. the 1996 AmLaw 100, and see how The American Lawyer's premises hold up.  1996 is, I believe, a particularly suitable year for a comparison to today.  It was, most importantly, a "normal" year:  It was well past the recession of the early '90's, and well before the tech/dot-com boom of the late '90's.  A time, in other words, much like today, with healthy but not insane growth.

What do the numbers tell us?  (Hint:  A very different story than The American Lawyer's take.)  Here's the raw material:

AmLaw 100	1996	2006
Revenue ($$ millions)	$16,214	$51,200
Headcount (lawyers)	38,156	70,161
Average PPP	$467,200	$1,060,000
Inflation adjustment (per Bureau of Labor Statistics)	$16.21	$20.77 (equivalent)

And here's what it all means:

• Absolute Growth in Revenue, 1996 — 2006:  +215.8%
• Inflation-Adjusted Growth in Revenue:  +146.5%
• Headcount Growth:  +83.9%
• Absolute Growth in Profits per Partner:  +126.9%
• Inflation-Adjusted Growth in PPP:  +99.1%
• Absolute Growth in Revenue per Lawyer:  +171.7%
• Inflation-Adjusted Growth in Revenue per Lawyer:  +134.1%

Now here's how The American Lawyer reads the numbers (to be sure, their comparisons focus on today vs. 1986 and today vs. 2004 and 2003) (emphasis supplied):

"In the U.S.'s 100 highest-grossing firms, in other words, additional lawyers are slowing the rate of revenue growth. Firms have continued to improve profitability by cutting costs, boosting efficiency, and making equity partnership more exclusive, but RPL remains the barometer of The Am Law 100's overall health. And, over the long term, RPL expansion is not keeping pace with head count increases."

My reaction is:

• Ratio by which growth in RPL has exceeded growth in headcount (1996 — 2006):  171.7%/83.9% = 2.05.
• And using inflation-adjusted RPL:  134.1%/83.9% = 1.60

"Additional lawyers are slowing the rate of revenue growth?"  How do you read these numbers? 

Better yet, to what metrics are you managing your firm?  I'd bet a healthy cross-section of executives across corporate America would look on these figures green with envy.