The #MeToo movement has helped to make nondisclosure and nondisparagement agreements notorious. Foisted on workers, designed in part to protect corporate trade secrets, they perpetuate harmful conduct by preventing people who survived abuse, harassment, discrimination and other harrowing experiences at work from telling their stories. The Weinstein Co. — co-founded by convicted sex offender Harvey Weinstein — famously wielded NDAs with a heavy hand, for example. Although there are countless anecdotal examples in which workers say they kept silent because of NDAs, we’ve lacked a broader understanding of how such agreements shape the sharing (or withholding) of information about workers’ experiences.

A study we conducted will help remedy that situation: We explored how online reviews of companies by current and former employees changed after three states — California, New Jersey and Illinois — passed laws prohibiting firms from using NDAs imposed as a condition of employment to conceal unlawful activity. (The laws we studied do not cover NDAs signed in settlement or severance agreements). These “NDA-narrowing” laws also protect whistleblowers. What emerged from the study was the first systematic, causal evidence that broad NDAs restrict the dissemination of negative information about companies. They thereby protect employers with harmful employment practices, prevent job seekers from learning valued information about potential jobs, and make it harder for the employers with the best workplaces to stand out.

We focused on the popular website Glassdoor — comparing how reviews changed in the states that passed NDA-narrowing laws with “control” states that did not pass such laws. After these laws passed, we found that ratings of firms (on a five-star scale) became 5 percent more negative (with the number of one-star reviews growing by 16 percent). Workers were also more likely to share specific negative information about companies: The “cons” section of reviews grew in length by 7.5 percent.

After the narrow-NDA laws were passed, workers who left negative reviews were about 20 percent more likely to share their job title — an important development, because information that is not anonymous is more credible and, therefore, more valuable to the job seeker. Reviews that specifically mention harassment of some kind — identified by searches for such words as “abuse,” “assault,” “sexual” and “victim” — increased by 22 percent. (Our interpretation of the data is that the narrow-NDA laws freed current and former workers to speak of past incidents more freely, not that incidents of harassment became more common after the laws were passed.)

The gap in ratings between the highest- and lowest-rated firms — in a given state and a given industry — also grew. That means that, in the narrow-NDA states, good companies were better able to distinguish themselves from those with bad practices. Greater differentiation between firms’ reputations should, in theory, inspire the worse-rated companies to improve — and there’s already evidence companies introduce reforms in response to Glassdoor reviews.

In light of the evidence we present, policymakers should reconsider their tolerance of broad NDAs and nondisparagement clauses. While narrow NDAs have legitimate purposes, such as protecting trade secrets, allowing employers to use NDAs to prevent workers from later speaking honestly about their work experiences has little to no economic benefit to society. Indeed, it makes the job market more inefficient, by concealing bad behavior.

Some argue that broad NDAs are justified because workers receive compensation for signing them. In some high-stakes settlement or severance arrangements (again, not the focus of the laws we studied), this may be true — though many later regret giving up their voice. (California’s legislature recently passed a law that would place tight limits on NDAs stemming from settlements; it’s awaiting Gov. Gavin Newsom’s signature.) But when required to sign an NDA as a condition of hire, workers receive little, if any, compensation (in the form of higher salaries), studies find.

For too long, the silence produced by overly broad NDAs has shielded workplace harms. California, New Jersey and Illinois have led the way in changing that situation. Now that we know that narrowing NDAs can help workers make better decisions, other states — and perhaps Congress — should consider narrowing NDAs signed as a condition of employment. That would allow even more workers to speak honestly about their worst experiences in the workplace.