Cops, Teachers, and the Art of the Impossible:Explaining the lack of diffusion of impossiblejob innovations

Document Type


Publication Date



In their now classic Impossible Jobs in Public Management, Hargrove and Glidewell (1990) argue that public agencies with limited legitimacy, high conflict, low professional authority, and weak "agency myths" have essentially impossible jobs. Leaders of such agencies can do little more than "cope," which is also a theme of James Q. Wilson (1989), among others. Yet in the years since publication of Impossible Jobs, one such position, that of police commissioner has proven possible. Over a sustained 17-year period, the New York City Police Department has achieved dramatic reductions in crime with relatively few political repercussions, as described by Kelling and Sousa (2001). A second impossible job discussed by Wilson and also by Frederick Hess (1999), city school superintendent, has also proven possible, with Houston and Edmonton having considerable academic success educating disadvantaged children. In addition, Atlanta and Pittsburgh enjoyed significant success in elementary schooling, though the gains were short-lived for reasons we will describe. More recently, under Michelle Rhee, Washington D.C. schools have made the most dramatic gains among city school systems. These successes in urban crime control and public schooling have not been widely copied. Accordingly, we argue that the real conundrum of impossible jobs is why agency leaders fail to copy successful innovations. Building on the work of Teodoro (2009), we will discuss how the relative illegitimacy of clients and inflexibility of personnel systems combine with the professional norms, job mobility and progressive ambition of agency leaders to limit the diffusion of innovations in law enforcement and schooling. We will conclude with ideas about how to overcome these barriers.


Working Draft - August 2010, Paper prepared for the 2010 Annual Meetings of the American Political Science Association