General counsel deserve to have good, reliable, and useful benchmark figures. Many people try to supply that need. All of them face some systemic unknowns about the representativeness of the law departments that participate in their surveys.
Centralized departments: For example, data may be less common from decentralized legal functions as well as less complete (See my post of Feb. 16, 2011: perhaps not full staff or spending figures.).
Sizable departments: Participants may tend toward the larger size since they are pursued more (See my post of July 31, 2011: possible bias toward size.).
Well-managed departments: Well run departments might disproportionately participate to confirm their beliefs and boast their prowess (See my post of Feb. 19, 2010: several forces that might influence survey representativeness.).
Listed companies: Privately-held companies are probably under-represented (See my post of Feb. 23, 2010: incomplete data from secretive companies; March 31, 2010: predominance of privately held companies in one survey; and Jan. 19, 2008: unknown degree of lower spending in non-publicly traded companies.).
English-speaking departments: Unless you speak a language fluently, you are more likely to get participants in your native tongue (See my post of June 13, 2010: bias toward those who respond to the English version of General Counsel Metrics global benchmark survey.).
Pressured departments: when besieged by pressures to cut costs, law departments might see support from comparative metrics (See my post of Feb. 19, 2010: possible tilt toward better-run departments.).