This is from Erin Metz McDonnell:

Within seemingly weak states, exceptionally effective subunits lie hidden. These high- performing niches exhibit organizational characteristics distinct from poor-performing peer organizations, but also distinct from high-functioning organizations in Western countries. This article develops the concept of interstitial bureaucracy to explain how and why unusually high-performing state organizations in developing countries invert canonical features of Weberian bureaucracy. Interstices are distinct-yet-embedded subsystems characterized by practices inconsistent with those of the dominant institution. This interstitial position poses particular challenges and requires unique solutions. Interstices cluster together scarce proto- bureaucratic resources to cultivate durable distinction from the status quo, while managing disruptions arising from interdependencies with the wider neopatrimonial field. I propose a framework for how bureaucratic interstices respond to those challenges, generalizing from organizational comparisons within the Ghanaian state and abbreviated historical comparison cases from the nineteenth-century United States, early-twentieth-century China, mid- twentieth-century Kenya, and early-twenty-first-century Nigeria.

…… Monolithically dysfunctional administrations are the exception, not the rule— albeit the exception that has long captured popular and academic attention (Evans 1989; Helman and Ratner 1992). Instead, many states regarded as uniformly ineffectual have great internal variation, with agencies spanning a continuum from ineffectual quagmires to competently achieving organizational man- dates in the public interest. These state “leviathans” are patch-worked: they are cobbled together from scarce available resources, with organizational diversity sewn loosely together into the semblance of unity. In such states, adapted Weberian-style bureaucracy exists in interstices—niches within predominantly neopatrimonial administrations.