Finally, and more importantly, the majority of existing studies on stateimposed local TELs have simply ignored the endogeneity of the policy and treated the policy as an exogenous event that happened autonomously or independently of the fiscal outcomes (e.g., Mullins and Joyce 1996; Johnston, Pagano, and Russo 2000; Mullins 2004). However, in reality, some unobserved voter attributes may affect both the enactment of TELs and the fiscal outcomes such as spending or revenue levels or different financing methods. For instance, voters in states with higher local taxes may be more apt to vote for a TEL and thus a positive relationship between TELs and taxes can be found. Yet, voters may simply change their preferences and want a lower level of taxes by voting for a TEL, thereby resulting in a negative correlation between TELs and taxes. Overlooking the endogeneity of TELs may cause biased and inconsistent estimates on the policy’s fiscal impacts (Poterba and Rueben 1995; Rueben 1996; Shadbegian 1999). To date, only a few studies such as Shadbegian’s (1998, 1999) seminal work and Dye, McGuire, and McMillen’s (2005) distinctive study have tried to correct the endogeneity problem of state-imposed TELs. The present study builds upon these earlier studies and aims to provide additional insight into the impact of TELs using an instrumental variable approach.