Why the Risk Lands First on Junior Women and Fragile Startups
The subject here is the uneven hiring response of venture-backed startups after harassment-related NDA reforms. The specific question is why a rule that applies broadly across a state produces its sharpest reduction among junior women and in small, male-dominated firms. Those differences help distinguish a firm changing its demand for labor from workers changing their willingness to join, while also showing why “internal capacity” is a useful explanation but not a directly measured variable.
Seniority reveals the power relationship better than skill does
A first possibility is that startups simply cut hiring among workers who are easier to replace. If so, the effect should be strongest by skill: perhaps firms would continue recruiting highly educated or technically scarce women while reducing hiring for less specialized positions. The study does not find a clean version of that story. Female hiring falls among both high-skilled and low-skilled workers, and the difference between the two groups is not statistically significant.
The sharper divide is seniority. The estimated reduction is roughly 10% for junior female hires, while the estimate for senior female hires is smaller and statistically insignificant. The researchers measure seniority in two ways: by inferred age and by job title. Age is available for only about half of the relevant workers, so those measures should not be treated as interchangeable. Both are intended to capture a position in the organization’s hierarchy, not a person’s value or competence.
That distinction matters because harassment risk is not distributed randomly across an org chart. A junior worker is more likely to depend on a senior colleague for assignments, evaluation, mentorship, and continued employment. If a firm’s reporting system is weak, the person with authority over a junior worker may also be close to the person who receives or suppresses a complaint. The legal reform changes the consequences of that relationship: if a covered dispute cannot be reliably contained by a predispute NDA, a failure of supervision may become visible outside the firm.
The result should not be read as saying that junior women are inherently risky hires. The risk in the study’s proposed mechanism comes from a power structure and from the firm’s ability to manage it. A firm that cannot prevent abuse or respond credibly may associate the most exposed relationship with greater expected liability. Reducing hiring at that point is a form of defensive demand response. It is also a way of transferring the cost of weak governance onto the people the policy was meant to protect.
Nor does the subgroup pattern prove that a hiring manager consciously thought in these terms. The study observes the distribution of hires, not the reasoning behind a rejected application. Junior women may also react more strongly to information about harassment and avoid startups in reforming states. The seniority result is therefore evidence about which mechanism is more plausible, not a direct measurement of intent.
It is useful to separate two claims that can otherwise become tangled. The first is descriptive: after reform, the fall in female hiring is concentrated among junior workers. The second is interpretive: this concentration is consistent with firms reducing demand where the supervisor-subordinate relationship creates more perceived exposure. The first is an observed pattern in the study. The second depends on assumptions about how firms and workers respond to the legal change.
The same distinction explains why skill and seniority should not be collapsed into one category. Education and occupation can describe what a person knows or does. Seniority describes where a person sits in a chain of authority. A highly educated early-career engineer may be both highly skilled and structurally junior. A less technical manager may have more power over a subordinate.