There was an interesting piece in Science last week about the long-running debates around the meaning and purpose of the “Broader Impacts” requirement on NSF grant applications. The article does a good job of articulating the complaints of scientists doing “basic” science that seems to far removed from “applications” to have anything meaningful to say about societal impacts. But it doesn’t ask whether such requirements over time force applicants to actually change their practices. I am sympathetic to arguments that requirements like this result in mere lip service—I have a strong dislike for policies that obfuscate the real work that justice requires. Similarly, as discussed at a conference at UCSF this summer, Leigh Star objected to affirmative action on the basis that institutions are able to develop a veneer of diversity without actually doing the work needed to maintain it, such as valuing the mentoring of minority students in tenure reviews. If the Broader Impact statements mostly produce vague and disingenuous claims about the societal importance of research, then they might actually do more harm than good because the applicants will come to view social justice as a burden or hurdle. Perhaps the goals of the broader impact statements would be better served with nationwide requirements that all institutions actually pursue and maintain a breadth of policies meant to develop justice-oriented research, rather than tacking goals artificially on to the end of NSF proposals. Yet it is hard to determine what the right answer is without a comprehensive look at whether and how the practices of scientists change over time when they need to consider these benefits within their research applications.