In this paper I use the Bush Administration's Program Assessment Rating Tool (PART) scores—a numerical measure of federal program performance—to analyze the relationship between political appointees and management. I find that federal programs administered by politically appointed bureau chiefs get systematically lower PART evaluations than programs run by bureau chiefs drawn from the civil service. I find that career managers have more direct bureau experience and longer tenures, and these characteristics are significantly related to performance. Political appointees have higher education levels, more private or not-for-profit management experience, and more varied work experience than careerists but these characteristics are uncorrelated with performance. I conclude that reducing the number of appointees or increased sensitivity to appointee selection based upon certain background characteristics could improve federal bureau management.