157 U. Pa. L. Rev. 833 (2008)
For the first time in American history, all nine Justices of the U.S. Supreme Court came to their positions directly from U.S. courts of appeals. As new vacancies arise in the coming years, should the President continue to look to the circuits for Supreme Court nominees? Commentators disagree on the answer. Those who support the current practice claim that the Senate is more likely to confirm nominees with judicial experience. Proponents also argue that former federal judges are more likely to reach decisions based on precedent rather than on their own ideological values. Those opposed to current practice point to the costs of elevating federal judges. Among the most pernicious may be "circuit effects," or the possibility that former U.S. courts of appeals judges are predisposed toward affirming decisions of the institutions they just left their respective federal circuits. We enter this debate not by rehashing the existing arguments, but by exploring them empirically. From our analyses, a clear conclusion emerges: the benefits of drawing Supreme Court Justices from the circuits are, at best, overstated, while the costs are, at a minimum, understated. Indeed, the data reveal a strong predilection on the part of Justices with federal judicial experience to rule in favor of their respective home court. For some, the attachment is so strong that they are twice as likely to affirm decisions coming from their former circuit as decisions coming from all others. Even more striking is the advantage now enjoyed by the U.S. Court of Appeals for the District of Columbia the former home of four sitting Supreme Court Justices. An obvious antidote is for the President to end the practice of appointing Supreme Court Justices from the circuits, & instead turn to the nation's law schools, law firms, legislatures, executives, & state courts. A less obvious, though no less plausible, remedy is for the President to select nominees from circuits underrepresented on the Court. Adapted from the source document.