This study initiates a methodological critique of the state‐level immigration policy literature through the lens of the racial threat and group power perspectives. First, I highlight the conceptual problems related to the application of such theories to legislature‐level data analysis. Next, I demonstrate the methodological and analytical problems that raise concerns about ad hoc theorizing in this field. Using counts of hostile and welcoming immigration legislation (2005–2011), I demonstrate that outgroup size measures correlate positively with both dependent variables while measures of population growth rate yield null results unless used on their own. These results suggest that the use of legislature‐level models with demographic indicators does not allow us to gain a clear understanding of whether and how population dynamics influence immigration policymaking. Based on these findings, I recommend that when using demographic indicators as key explanatory variables, researchers provide evidence of result consistency across multiple model specifications and also test the models with both hostile and inclusive policy variables. Such protocols would help avoid ad hoc theorizing.