Published benchmarks may be tougher than what actually prevails because of selection bias

The general counsel of inefficient law departments don’t want to provide comparative data that confirm they are operating inefficiently. They are probably less likely to choose to participate in benchmark surveys than their fitter counterparts. Not that anyone sees any specific company’s numbers, because the data are aggregated and normalized. Even so, if companies with less flattering benchmark figures systematically don’t take part, perhaps claiming ill-advised concerns about confidentiality, their numbers are not in the data and the benchmark results conceal mediocrity. Published benchmarks, therefore, may be better than what actually and typically prevail.

This is the mischief of self-selection bias. Well-managed departments want to prove their quality so they take part disproportionately in benchmark surveys. What does self-selection cause? Some foundational findings, like total legal spending as a proportion of corporate revenue, may be higher than the benchmarkers state. If only the lean ones weigh themselves, the world is fatter than the resulting medians and averages. There is no mechanism yet by which a large group of law departments must provide figures that test this hypothesis of self selection.

The best we can hope for until then is to have trade groups reach out and urge wider participation among their members. Or, to have benchmarking be seen as simply what everyone does. Sheer numbers of participants means the results are likely to dig deeper and be more representative.

We welcome comments

Your email address will not be published. Required fields are marked *