The Inside Response: Women Move Up as the Pipeline Narrows

The subject of this chapter is what treated startups do after harassment-related NDA reforms change the hiring calculus. The focus moves from the external question—who enters the firm—to the internal transitions that follow: who is promoted, who leaves management, whether women returning to work gain access, and how financing changes. Tracking those transitions supports a more careful judgment than the headline alone: a firm can become more accountable for people already inside while still narrowing the pipeline from outside.

From a smaller pipeline to a different leadership hierarchy

The NBER study reports an increase in the female share of promotions into manager- and director-level positions after the reforms. The estimated change is about 3 percentage points and is statistically significant at the 10% level. A similar-sized estimate appears for promotions into junior or associate roles, but that result is not statistically significant. The concentration of the clearer result at the management level matters because managers occupy the part of the hierarchy where authority over subordinates, supervision, and exposure to harassment-related disputes meet.

This internal movement can coexist with fewer female hires for a simple reason: promotion and recruitment draw from different pools. A startup may promote a woman who is already employed while reducing the number of women it recruits into entry-level or early-career roles. The first decision changes authority among existing workers. The second changes who gets the chance to become part of the organization. Treating the two as one measure of “female employment” would hide the trade-off.

The paper’s interpretation is that elevating women into management can reduce the firm’s exposure to male-on-female harassment in supervisor-subordinate relationships and weaken the glass ceiling. But that interpretation should be stated as a risk-mitigation explanation, not as proof that every promotion was made for legal reasons or that every promoted woman replaced a problematic manager. Aggregate promotion shares cannot reveal the deliberation behind an individual decision.

The other side of the transition is managerial departure. After reform, the male share of departing managers rises by 2.7 percentage points. The study does not find a corresponding effect among non-managerial male workers. That specificity is consistent with a change centered on authority and supervisory relationships rather than a general male exodus from the startup workforce.

What happens to departing managers afterward adds information, but not certainty. Their probability of moving to a lower-salary job rises by 3.6 percentage points; the probability of moving to a lower-seniority position rises by 3.9 points; and the probability of experiencing an unemployment spell longer than the sample median rises by 5.1 points. If managers were simply leaving voluntarily for better opportunities, improved subsequent outcomes would be the more natural expectation. The reported deterioration therefore suggests that at least some separations may have been involuntary.

“May” is doing important work in that sentence. The data do not identify a harasser, record an employer’s reason for termination, or show that any particular departure reduced future misconduct. A firm could remove a manager it viewed as a legal or reputational risk. A male manager could also withdraw because closer interaction with female subordinates now feels personally risky, producing a broader chilling effect rather than a targeted accountability response. The same employment transition can look similar in the data under those different mechanisms.

That ambiguity is not a reason to ignore the result. It tells us what kind of evidence the reform generates.