This is a very interesting article because it avoids the usual pitfalls of demography-based arguments. ("Group G is underrepresented in activity A. Therefore we

This is a very interesting article because it avoids the usual pitfalls of demography-based arguments. ("Group G is underrepresented in activity A. Therefore we should completely overhaul A with policy P, which appears to be pro-G. If you have any concerns about P it can only mean you're an anti-G bigot.") The article instead asks a novel question and leaves solutions open. How do tasks get divided among peer groups? It seems like people who are vulnerable to peer pressure (girls in engineering, in the examples here) are likely to get booted out of tasks they enjoy and relegated to routine chores instead. It's a testable hypothesis, it opens a new perspective, and if proven true it lends itself to very reasonable remedies for both employers (e.g. formalize the task assignment process) and the affected engineers (watch out and prepare to stand up to it if needed).