With the rise of misinformation epidemic, this study aims to empirically investigate the consequences of an online commenting platform's activity-capping policy on abusers' and regular users' activities. Utilizing a quasi-experimental setting, we find that restrictive policies not only curtail the activity of the abusers but also promote the activity of regular users. Results show that the policy has an asymmetric effect on abusers and regular users- while it effectively reduces the actions of the malicious users by 1.8%, it promotes the activities of the regular users by 2.2%. To better understand the behavioral change of the regular users, we draw from the rational economic perspective of voting decisions and provide initial evidence that such policy measures reinforce the subjective probability of being influential on the outcome. This study will provide valuable implications to managers and policymakers to estimate the consequences of and to combat against malicious behaviors and to promote free speech in online platforms.