2021-10-16 | PhD committees in political science and recruitment

In this paper, we discussed which type of network structure is the most efficient. Porous (a.k.a. structural holes) or dense (a.k.a. closure) ones? We made the claim that those two types are not necessary contradictory. They provide benefits at different levels.

Porous ones (with little relational redundancy) empower an individual within the group. Dense ones (with strong relational redundancy) empower the group in its competition with other groups.

In order to test these ideas, we used 1990s French PhD Committees in Political Science as our network structure and academic recruitment as our outcome variable.

This figure (and a regression) illustrates the individual dimension. On the center right of the figure, supervisors with multiple and diversified connections had their PhD students hired (cf. last label number). Those on the left, with repeated and cohesive ties don't.

Image

To illustrate the collective dimension, we showed the  very cohesive network structure of Toulouse (in the 90s) which enabled them to be strongly united and defend their own PhD students when they had the power to recruit them. Consequence: strong academic inbreeding there.

In contrast, Strasbourg displayed in the 90s little cohesion and we claimed it explained why they didn't engage at all in academic inbreeding.

Translation remains a bit rusty and would need more Englishing. Sentences would benefit more shortening and reordering (too much appositions). Still after 30 seconds DeepL processing and two hours of editing, the balance seems fair and I hope readable.