Some states implement voting laws that make it harder to vote. Today, despite evidence that voter impersonation is rare, 34 states have implemented some form of voter ID. I ask: 1) Where do restrictive voting policies come from? 2) Why do state legislators pass restrictive voting policy? What conditions make legislators more likely to sponsor and support these policies? Few studies examine what factors contribute to state legislator support for restrictive voting requirements. I investigate how demographic threat impacts state legislator support for voter identification legislation. I argue that voter identification laws are elite-driven devices that legislators use to retain political power when changing demographics disfavor them and pose an existential threat to their electoral viability, particularly when it is unfeasible for legislators to successfully win support from racial minority voters. The restrictive voting laws prior to the civil rights movement were explicitly racially discriminatory, but in order to overcome legal challenges restrictive voting laws today are written to be race neutral and to produce a racially discriminatory effect. This study takes a holistic approach, examining legislator intent behind voter identification laws, by using historical context around restrictive voting laws to establish the link between the restrictive voting laws of the past and those of today in the literature. First, I examine archived sources to understand how voter identification laws developed from other discriminatory tests and devices. Findings indicate that early voter identification laws were implemented in places where large nonwhite populations enjoyed strong political organization and power, and modern voter identification laws are adopted similar contexts. Next, I examine why state legislators vote for and sponsor these laws. I find evidence that state legislators respond to demographic threat with restrictive voter identification legislation. Many scholars examine voter identification laws through voter fraud and partisanship, but I argue that this ignores 150 years of targeted voter suppression against Black people, Latinos, and immigrants. These findings have broad implications for understanding how state legislators respond to demographic change.