Workplace discrimination tracks how easily a worker can quit, a 32-country study finds

Some arithmetic first. The reported share works out to at least about 3,680 people carrying the outcome variable ^[13], and each country contributes roughly 625 respondents on average ^[14]. That is workable for a pooled correlation and thin for the claim a reader will want to make, which is that the pattern holds in their own labor market. The phys.org account does not say whether the comparison is drawn within countries or across them ^[12].

Measurement is where I keep circling back. Discrimination here is what a worker reports having experienced ^[1], and at least two processes would produce the same association. Van Hoorn's own framing contains both: mobility as a shield with the current employer, and mobility as an escape route to a new job ^[15]. Under the shield reading, an employer who can see a valued worker leaving steps in and conditions improve ^[16]. Under the escape reading, the mobile worker exits early, the five-year tally is shorter, and nothing about the employer changes. Both fit a bargaining-power account, but they differ in who benefits: deterrence helps the colleagues who stay, while exit moves the same treatment onto whoever cannot follow.

The thing this doesn't tell you is whether any employer did anything. As summarized, the study observes worker reports rather than managerial behavior ^[12]. And the traits that make a worker easy to replace can raise exposure directly, with no employer weighing replacement cost at all, which is confounding the summary gives no way to bound ^[12].

What would settle it is a change in mobility the worker did not choose. Van Hoorn expects better trains and buses to put more employers within reach ^[6], and he notes that a dense local supply of similar vacancies makes leaving more likely ^[5]. If reported discrimination falls among affected workers after nearby options multiply, with no reason to think local attitudes moved, the bargaining channel is doing real work.

The policy reading is the interesting part regardless. Van Hoorn is candid that anti-discrimination law is weakest where the behavior is most common, in the small everyday cases ^[8], and he points out that firms differ in how much they discriminate ^[9], which is what gives an exit route any value at all. He also says, without hedging, that mobility does not replace work on root causes ^[10]. That conditioning matters, because an escape route is a remedy the worker has to spend, and spending it leaves the employer's conduct where it was.