Women's work friendships grow by introduction, men's by selection, Durham study reports

Two ways of acquiring a colleague are being compared here, and they have different geometry. A tie that arrives through a mutual contact lands inside a group whose members already know each other, and the paper reports that these friend-of-a-friend introductions were more likely to be to other women ^[2]. A tie formed one to one, with nobody brokering it, can land anywhere, including outside the cluster the new hire started in ^[3]. The researchers found men's selective ties were especially common where the relationship was perceived to support a personal or professional goal ^[6].

That is the part worth sitting with, because nothing in it requires anyone to prefer men or to intend exclusion. One process builds density. The other builds reach.

It also explains how the study's two headline findings can both hold. Women ended up with the larger and more interconnected networks ^[1], and the authors nevertheless argue those networks supply fewer routes to influential contacts, resources and career-enhancing connections, while men's more strategic ties are likelier to yield professional advantage and preferential treatment ^[13]. Both are true if what promotion responds to is not how many friends you have but how many separate places your information comes from. Nine people who all know each other are, as an information source, much closer to one than to nine. The team extends the argument to the gender pay gap ^[14], which is a longer leap than the data described in the write-up can carry on its own.

On the measurement: almost 150 employees, four survey waves, the first ten months in the organisation ^[5]. That averages one observation roughly every two and a half months ^[9], which is a reasonable cadence for catching cluster formation as it happens rather than reconstructing it afterwards. The press account reports no effect sizes and no gender composition for the sample ^[10], so how much larger and how much more same-gender these networks were remains unread until the paper itself is. Note also that the summary names Andrew Parker of Durham as lead, with colleagues at William & Mary's Raymond A. Mason School of Business and Universita della Svizzera Italiana ^[4], while the journal citation is first-authored by Inga Carboni ^[8], a discrepancy in emphasis rather than substance ^[11].

The practical reading is narrower than the researchers' own recommendation of broader networking opportunities ^[7]. Closure needs a broker. Someone has to be the third person who makes the introduction, and in a new hire's first months that person is usually assigned by a manager or an onboarding schedule. If the tie-formation mechanism is what produces the difference, then the roster of who introduces whom to whom is where an employer either reproduces the pattern or interrupts it, and it is one of the few things about informal networks that a company actually controls.