Dominant groups worldwide have historically asserted power by constructing in public spaces monuments that glorify their narrative, vis-à-vis their opponents’. How do divisive public símbols affect the location choices of those who oppose them? I investigate this historically and today, focusing on Confederate monuments in the US South – erected by southern whites in the early 20th century and opposed by Black Americans due to their connection to slavery. Historically, I show that southern counties with monuments saw a sharp decline in the Black share of the population – driven by out-migration – following their construction. However, monuments themselves are outcomes of underlying ideological shifts, making causal claims problematic. I thus construct an instrument for the stock of Confederate monuments based on transportation costs to a quasi-monopolist producer and the years in which it was in business. The IV analysis confirms that monuments caused a substantial reduction of the Black share of the population. I complement the historical analysis with an online experiment to assess whether monuments still influence migration choices today. I randomize Confederate monuments in the visual depiction of hypothetical destination cities and ask respondents to consider job offers there. Black respondents request higher reservation wages and are significantly less likely to accept offers.