And here I was hoping that someone had scraped LinkedIn for people's employment histories, then built a state machine/markov decision process-esque model of the tech industry, and quantified how much the collusion depressed inter-company movement.
Considering LinkedIn is suing[1] left and right for people scraping LinkedIn profile data.. Any analysis would most likely be released anonymously.
That would be fairly damning evidence. Compare number of employee transfers between allegedly colluding companies to average number of transfers for compared companies and if collusion is real numbers will tell.
I'm not sure that would elide the desired conclusion, or at least colluding companies could explain the lack of movement as company desirability, stability, and fair compensation.