Multilevel Methods in the Study of Bureaucracy

Carolyn J. Heinrich and Carolyn J. Hill

in The Oxford Handbook of American Bureaucracy

Published in print October 2010 | ISBN: 9780199238958
Published online January 2011 | | DOI:

Series: Oxford Handbooks of American Politics

Multilevel Methods in the Study of Bureaucracy

More Like This

Show all results sharing these subjects:

  • Politics
  • US Politics
  • Political Methodology


Show Summary Details


This article first presents a brief overview of the major types of multilevel statistical models. It reviews two major approaches to multilevel modeling available to researchers today: cross-sectional and longitudinal. It describes the case for their theoretical and conceptual value in the study of American bureaucracy and their potential technical advantages over more conventional approaches to statistical estimation of relationships in governmental systems. Next, the methodological critiques and limitations of multilevel modeling techniques are addressed. In particular, three of the major methodological critiques of multilevel models are explained: the causal inference problem, the argument that these models offer no net advantage, and cross-discipline transferability critiques. Finally, the article determines some of the most promising avenues for the future application of multilevel models. These include their use to promote more effective accountability systems by emphasizing performance analysis, to improve modeling of interorganizational relationships, and to investigate cross-level relationships.

Keywords: multilevel statistical models; American bureaucracy; multilevel modeling; causal inference problem; cross-discipline transferability

Article.  9314 words. 

Subjects: Politics ; US Politics ; Political Methodology

Full text: subscription required

How to subscribe Recommend to my Librarian

Buy this work at Oxford University Press »

Users without a subscription are not able to see the full content. Please, subscribe or login to access all content.