The Institution Behind the Result

The comparison question is now sharper: if Brazilian judges recover observed output after childbirth and almost all remain employed, is that because they are unusually capable women, or because their jobs make recovery and retention possible? This chapter compares judges with self-employed labor lawyers, permanent public-sector workers, and private-sector workers who resemble judges along important dimensions. The practical judgment is to separate the portability of the output result from the portability of the retention result: a worker may recover measured activity without enjoying the institutional protection that keeps them in a job.

That distinction matters because “the public sector” and “high income” are not single explanations. Judges have autonomy, but so do some self-employed lawyers. Judges have high earnings, but public-sector employment stability also appears among lower-income workers. Private-sector comparison groups can share education, family background, occupation, or income with judges while facing different employer incentives. Looking across these groups helps identify which differences are most consistent with the outcomes, while preserving the fact that these are carefully chosen comparisons rather than one experiment that isolates every mechanism.

Output can recover without public-sector protection

The first comparison changes the employment environment while retaining some occupational similarity. Self-employed labor lawyers work in the same broad legal world as judges and have substantial control over their activities, but they do not receive the judges' employment protection or statutory parental leave. Their average income is also much lower: the paper reports that judges earn about 3.53 times as much as these lawyers. If public protection were necessary for a parent to recover observed output, the lawyers should show a lasting decline after childbirth. If high income alone were the key, the lower-paid lawyers should fare very differently from judges.

The evidence does not fit either simple prediction. The study examines nearly 300,000 self-employed labor lawyers and measures their output through new cases filed and cases won. Mothers show a temporary reduction around childbirth, especially in new cases filed, but activity recovers within a few months. The post-leave-period estimates do not show a lasting effect on cases filed or won. Additional outcomes—including sentence amounts, case duration, appeals, and reversals—also do not show significant persistent changes for either mothers or fathers.

This comparison is useful because it separates two kinds of portability. Lawyers are not protected employees, so their experience speaks to whether legal work can recover after a parenthood-related interruption without a guaranteed return to a public job. But they are self-employed, which means the decision to pause, resume, or reorganize work is not the same as a firm's decision to retain an employee. Their case allocation is also not random. The comparison can therefore show that output recovery is not unique to judges, while remaining unable to identify how an ordinary private employer would respond to a returning mother.

The temporary lawyer decline is also informative in its own way. Because these lawyers do not have statutory parental leave, the dip cannot simply be read as the mechanical recording of a formal absence with a fixed institutional duration. It is consistent with a private adjustment in activity around childbirth. The subsequent recovery suggests that an interruption can resolve even when the worker lacks the judges' formal protection, but it does not reveal whether the cost was absorbed by the lawyer, by clients, by household support, or by a change in the timing of cases.