Political advertising on social media exposes how algorithmic targeting intersects with nationalist discourse and demographic bias. By analyzing over two thousand migration-related campaigns from the Facebook Ads Library, this study uncovers how anti-immigration messages reach audiences differentiated by gender, age, and location, amplifying visibility among groups aligned with nationalist sentiment. The results reveal how algorithmic mediation reinforces cultural divides, showing that persuasion in digital politics depends as much on audience composition as on ideology.