Improving Education through Bureaucrat Autonomy in Bangladesh
How to cite
Lopez-Pena, Paula; Haque, Ashraf; Mobarak, Mushfiq; Wen, Jaya. 2018. "Improving Education through Bureaucrat Autonomy in Bangladesh." AEA RCT Registry. AEARCTR-0003513. https://doi.org/10.1257/rct.3513-1.1
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
A long-held view in the economics literature is that lack of autonomy and incentives to exert effort or take risks stymies efficiency and the adoption of new practices by bureaucrats (Mookherjee 1998, Bloch and Bugge 2013). And yet, evidence on how to design incentive schemes that effectively prompt government officials to improve the quality of education remains limited (Wild et al. 2012). We propose a novel experiment in which we cross-randomize training in evidence-based and promising educational interventions and two types of incentives - towards a final goal, and towards a set of intermediate goals - among Upazila Nirbahi Officers (sub-district level government officials) in Bangladesh. The study aims to provide new insights into effective incentive design by investigating what combination of autonomy and non-financial incentives is more effective in increasing policy innovation and improving student learning outcomes.
Experimental design
We randomly select Upazila Nirbahi Officers (UNOs) from 220 upazilas in Dhaka and Chittagong and allocate them into four experimental arms of 55 upazilas each: (1) a full autonomy arm, in which UNOs are paid incentives to improve a final outcome (namely, average student test scores in their jurisdiction); (2) a partial autonomy arm, in which, in addition to being paid a final incentive, UNOs receive a toolkit of evidence-based interventions for improving educational outcomes, are invited attend a two-day training in Dhaka, and receive follow-up phone calls to remind them of the contents of the toolkit; (3) a low autonomy arm, in which UNOs receive final incentives, are invited to attend the same training and receive the same follow-up calls as the partial autonomy group, but are also paid to take specific intermediate steps toward toolkit implementation; and (4) a pure control group, in which UNOs do not receive any incentives or training.
Design details
Cited by
No citation data available for this registration yet.