Published online by Cambridge University Press: 27 September 2018
Bureaucracies with field operations that cannot be easily supervised and monitored by managers are caught between two sources of dysfunction that may harm performance. The first source of dysfunction is straightforward: field workers can use operating slack and asymmetric information to their own advantage, thwarting an organization's objectives. The second source of dysfunction is often overlooked: attempts to limit workers’ autonomy may have deleterious effects, curbing agents’ ability to respond efficaciously to the environment. I find that the parliaments and executive boards to whom International Development Organizations (IDOs) are accountable differentially constrain IDO organizational autonomy, which in turn affects management's control of field agents. Tight management control of field agents has negative effects, particularly in more unpredictable environments. Attempts by politicians to constrain organizations in an effort to improve performance can sometimes be self-undermining, having net effects opposite those intended.
I thank the National Science Foundation Graduate Research Fellowship for its support under grant #DGE-1144152. Matt Andrews, Sam Asher, Nancy Birdsall, Mark Buntaine, Andreas Fuchs, Peter Hall, Conor Hartman, Bob Keohane, Chris Kilby, Steve Knack, Aart Kraay, Jane Mansbridge, Sheila Page, Woody Powell, Lant Pritchett, Simon Quinn, Steve Radelet, Tristan Reed, Alasdair Roberts, Bill Savedoff, Evan Schofer, Ryan Sheely, Beth Simmons, Vivek Srivastava, Martin Steinwand, Mike Tierney, Dustin Tingley, Eric Werker, Michael Woolcock, anonymous reviewers at IO, and many others have all provided helpful comments on earlier versions of these ideas and/or this work. Many thanks to Yi Yan, Smriti Sakhamuri, Grace Chao, Kyle Kessler, and the World Bank Archives staff (particularly Sherrine Thompson) for their research assistance. I thank the European Commission, the UK's Department for International Development, the Asian Development Bank, the Global Fund for AIDS, Tuberculosis, and Malaria, and the German Development Bank for providing data. Because World Bank data used in this analysis are publicly available the World Bank deserves perhaps the greatest thanks. Data for the Japanese International Cooperation Agency (JICA), the German Society for International Cooperation (GiZ), and the International Fund for Agricultural Development were assembled from individual project-completion reports by Odesk-contracted research assistants under my supervision, with the compiled data then sent back to the originating agency for comment and/or correction. GiZ was kind enough to respond with corrections, which were incorporated; these data were generated by me rather than by JICA and JICA is not responsible for them. The case studies would not have been possible without the hundreds of hours spent by a diverse group of interviewees listed in the online appendix who talked to me and answered odd questions, for which I am very grateful.