Articles Posted in Benchmarks

The ACC Docket, April 2011 at 14, states a finding from the 2010 Managing Outside Counsel Survey: “The median number of US law firms used by law departments during 2009 was 12.” A dozen seems too low. If a law departments has a couple of law suits during the year, a firm to review issues in each of a handful of specialty areas, local counsel, etc. it is be easy to hire more than a dozen, so to find that half of the companies in the survey use fewer needs some backup.

The summary offers one sentence more. “The average number of US law firms used by small law departments (2 or fewer lawyers) was five firms; medium law departments (3-10 lawyers) was 12 firms; and large law departments (more than 10 lawyers) was 60 firms.”

When the finding one year for large departments was 20 firms, I challenged it as too low (See my post of April 18, 2009: 2009 Serengeti survey gave 20 law firms.). This fuller set of data does not seem right to me either. The average of 12 means there had to have been mostly smallish law departments in the survey. Or perhaps if a department did not give a number the analysts treated it as zero firms? Did the phrasing of the question lead to an undercount?

Who, exactly, completes benchmark surveys?

I looked at the positions of the first 215 respondents to my General Counsel Metrics benchmark survey. 131 of them (65%) are general counsel. Another 30 (14%) are law department administrators and about the same number are direct reports to the General Counsel. In other words, nearly always the top legal officer personally completes the online survey or requests a direct report to complete it. The data, therefore, is more likely to be accurate and comprehensive.

A co-panelist with me recently mentioned that her law department regularly compares its fully-loaded hourly billing rate for lawyers with the average rate charged by outside counsel. The comparison has a fair amount of validity, but some faults can mar the match (See my post of June 29, 2009: insurance company’s calculation of savings based on fully loaded costs.).

The internal load needs to be full, that is, to include all the costs of internal lawyers (See my post of Aug. 27, 2008: fully-loaded cost per lawyer hour with 31 references; and March 9, 2009: fully-loaded with 7 more posts.).

The chargeable hours estimated or tracked for the internal legal group must be legitimate (See my post of May 21, 2009: internal chargeable hours with 12 references; and Jan. 5, 2011: internal time tracking with 9 references and one metaposts).

Aficionado that I am of metrics, I started slicing and dicing the revenue figures from the 215 law departments that have submitted data so far to the GCM global benchmark survey. Those companies reported $2.4 trillion of combined corporate revenue for 2010. (For those reporting in Euros, British Pounds and Canadian Dollars I used the average exchange rate to US dollars during the year.)

Poking around further, it turns out that 119 of those participants (55%) reported revenue greater than $1 billion. Moving up in size, 78 (32%) reported more than $3 billion and 56 (26%) broke the Fortune 500 barrier at more than $5 billion.

At the top: 34 of the participants so far exceed $10 billion. One out of seven loom that large; at the other end, about half the participants reported less than $1 billion in 2010 revenue.

General counsel hold their settlement data close to their vest. They fear repercussions in litigation if those amounts or terms leak out. Therefore, those who want to collect data on corporate settlements for benchmark purposes should dispense with direct inquiries yet try oblique ones.

Surveys could ask for the average amount paid over a three-year period. That smoothes out the peaks and conceals specifics. Or they might ask for a range or an estimate of only those amounts that have been publicly disclosed, such as fines paid to government agencies. Another way forward would be to ask for the aggregate minus the largest settlement, which is probably the most controversial or proprietary. Or the average – total settlements divided by cases settled during a year – would protect both the numerator and denominator.

It might also be easier to collect data on what law departments receive in settlements. Since amounts paid dwarf amounts collected for most corporations that would be provide meager insights.

In-house attorneys around the world work varying numbers of hours each year, at least if you give credence to the combined effect of paid leave and paid public holidays mandated in each country. Consider one flagrant contrast summarized by Eduardo Porter, The Price of Everything: Solving the mystery of why we pay what we do (Portfolio/Penguin 2011) at 75: Portuguese workers get a total of thirty-five paid days off a year whereas US workers have no mandatory paid days off.

Granting in-house lawyers comparable paid time off, with such significant variance from country to country, to use 1,800 chargeable hours per year as a baseline from US law departments almost certainly overstates chargeable hours elsewhere. That distortion, for example in calculations of fully loaded costs per lawyer hour, understates the costs overseas. If lawyers in France, Germany, and the UK for example, log fewer hours, of which hours we assume the same percentage of chargeable hours as their US counterparts, then their cost to their company rises as compared to US lawyers.

The vexing question remains of what is the appropriate formula to figure out the carrying cost of internal lawyers (See my post of July 26, 2009 #1: difference it makes if you change chargeable hours; May 21, 2009: internal chargeable hours with 12 references.).

For several reasons surveys don’t collect data on settlements paid by corporations. Often those amounts are kept confidential, and few within the defendant company know them. Other times settlement funds come from business units, sometimes more than one, and it may be difficult for the law department to assemble those scattered and embedded amounts.

At times, a settlement might be paid in more than one chunk. Most fundamentally, general counsel are loath to disclose to anyone the amounts, frequency or patterns of their settlements of litigation. Even aggregated numbers worry them. Disclosed, that sensitive knowledge they fear could give plaintiffs’ lawyers an advantage. Hence, practical, judicial and prudential obstacles thwart those who would like to produce settlement benchmarks.

PS. A scrupulous surveyor would ask not for gross settlements paid but net settlements paid.

A simple formula allows you to calculate the number of relevant documents missed by reviewers, or the precedent cases missed by researchers, or the improper expense records submitted on invoices, or the number of typos in a brief. Curious?

As explained in John D. Barrow, 100 Essential Things You Didn’t Know You Didn’t Know: Math Explains Your World (Norton 2008) at 11, you start by having two people do their best on a search for discovery documents, opinions, billing irregularities, or proofreading catches. You then calculate “the product of the number that each found that the other didn’t divided by the number they both found.” The closer the two people are to finding the same items, the more confident you can be that the unfound items are few; conversely, if each found many items not found by the other, the unfound items will be much more numerous. With this formula, you can at least size the gap.

On Monday my column appeared on InsideCounsel.com regarding three uses of metrics. Grandly, the uses are analysis, explanation and persuasion, the latter also known as rhetoric. It is useful to understand these powers inherent in credible numbers used skillfully. If you would like to find out more about this topic, click here for my three-powers column published on April 11, 2011.

My speculation is that the degree of competition in an industry strongly determines the industry’s legal intensity. Benchmarks such as those on personnel and dollars per unit of revenue will vary between industries because of greater or legal competition.

You can measure competition by the number of firms (both public and private) operating in the industry. For manufacturers, some researchers use the Census of Manufacturers as their data source. They measure manufacturing intensity through the ratio of the dollar value of manufacturing companies in the industry and the total dollar value of shipments in the segment or industry.

Once enough benchmark data is collected for industries, and better for segments within industries, it will be possible to test my hypothesis (See my post of Dec. 14, 2005: legal intensity of regulation; Nov. 24, 2007: intensity equals complexity, volume, and velocity; Aug. 13, 2008: R&D spending correlates to total legal spend; Feb. 24, 2009: intellectual capital of a company and legal intensity; March 2, 2009: industries vary in level of legal issues; Dec. 3, 2009: R&D intensity and legal costs; Aug. 3, 2010: little difference based on legal intensity in law firm assessments; Aug. 16, 2010: technological and competitive aspects of industries; Dec. 27, 2010: benchmark data on inter-industry differences; and April 3, 2011: litigation in relation to growth and competition.).