This paper empirically examines whether the act of deciding to support a political party
can impact partisan leanings years later. We use the discontinuity in the probability of being
registered to vote around the 18th birthday to look at the impact of registration after the 9/11/01
attacks on party of registration. We …rst show that 9/11 increased Republican registration by
approximately 2%. Surprisingly, these di¤erences in registration patterns fully persist over the
two year period from 2006 to 2008, even for a group of registrants who moved and changed their
registration address. We …nd full persistence for those registered in zip codes within two miles
of a four year university, suggesting that persistence is unlikely to be explained by lack of easy
access to or inability to process information. Instead, we suggest an interpretation of our …ndings
based upon either cognitive or social biases.