There have long been concerns regarding racial disproportionality in the U.S. child welfare system: Black children are represented in foster care systems at levels much higher than their numbers in the overall population. Calls for reform have grown louder in recent months, as the nationwide push to re-examine structural racism in institutions has reached the child welfare system. An increasingly popular reform seeks to reduce disproportionality by eliminating perceived implicit biases in the decisions of child welfare workers. This program, known as `"blind removals," works off of the following premise: if demographic information is unknown to professionals deciding whether or not to remove a child, then implicit biases will not impact foster care placement decisions. We conduct the first quantitative analysis of blind removals and derive two main findings. First, we show that the over-representation of Black children in most foster care systems is almost entirely driven by the fact that Black children are roughly two times more likely than White children to be investigated for child maltreatment to begin with. Conditional on initial rates of investigation, investigators remove White and Black children at similar rates. Specifically, we calculate that equalizing the removal probabilities of White and Black children would only reduce overall disproportionality by 4.5 percent. Thus, policies that target racial disparities in the removal decision have limited scope for impacting racial disproportionality in most foster care systems. Second, we find no evidence that blind removals had any effect on the already small racial disparities in the removal decision but they substantially increased time to removal. These findings yield an important insight for the multiple states and local child welfare agencies currently considering implementation of blind removals: the policy is not well-suited to reduce racial disproportionality in most foster care systems.