Strategies to counter misinformation have, so far, mostly focused on demand-side approaches, encouraging and equipping media users to avoid it, detect it, fact-check it or report it (Ecker et al. 2022, Jerit and Zhao 2020). Yet, since elites play a central role in the dissemination of misinformation (Mosleh and Rand 2022), many governments have begun to enact supply-side approaches assessing legal penalties to individuals that share misinformation. Brazil’s new regulations, introduced after the storming of the presidential palace in 2023, illustrate this trend. Since 2024, a key government body systematically monitors the online behavior of political candidadates, and energetically prosecutes offenders. In this survey-experiment, we explore whether informing candidates for office of these penalties (n=873) is effective in mitigating the flow of misinformation they endorse. We go beyond this and additionally test whether other messaging approaches can have the same effects. We show that informing candidates of penalties causes significant improvement in their accuracy discernment: they are able to identify and avoid misinformation at greater rates when they are informed of potential costs. However, we also find that exposure to these countermeasures reduces the overall amounts of content posted. Finally, we show gentler nudges may be equally as effective as raising penalties. These findings have broad implications for how governments and online platforms seek to durably address misinformation: strong supply-side approaches may reduce it, but do so at the cost of a general chilling effect on social media engagement.