How does physically witnessing a protest in a democratic society affect citizens in authoritarian societies? Existing data are unable to answer this question because of their difficulty in capturing witnesses, constructing meaningful comparison groups, and obtaining information on pre-protest political behaviors. Using a quasi-experiment design, I report a first causal estimation of the impact of 13 protests in Hong Kong from 2012-14 on witnesses from mainland China. I used geocoded posts from a Chinese social networking site to construct a panel of Chinese users who were otherwise similar but had visited Hong Kong at different time: treated users were physically close to one of the protests when it occurred while control users already left Hong Kong before the protests occurred and therefore could not witness them. I used difference-in-differences methods to estimate the causal effect of protests on userstextquoteright discussion intensity and issue of civic and political problems. First, The treated users published 40.39% more posts about civic and political problems after protest, relative to the change of the control users. This increase is robust under a replication study based on Taiwan, and three placebo tests. It remains significant within three months after the protests. Second, treated users discussed more issues about their daily lives. Rather than hoping to bringing similar changes to China, their discussion of democracy were mainly within the context of Hong Kong, and showed both support and opposition to democracy. The results suggest that witnessing protests driven by democratic claims lead citizens from authoritarian regimes to be more civically engaged.