Association for Public Policy Analysis & Management

Does engaging public managers in specific data-driven management exercises promote the cultural, attitudinal, and behavioral shifts required to support effective implementation of PerformanceStat-based managerial reforms?  Such reforms are extraordinarily difficult to implement in the public sector, as evident in the large number of such management initiatives at all levels of American governments that have either been abandoned or have failed to deliver their promised benefits.  A key challenge is the difficulty of moving organizations from basic measurement of metrics to active managerial decision-making informed by performance indicators.  At the same time studies suggest that inculcating among managers those routines that focus on organizational improvement can lead to greater analytic use of metrics.  

    We are engaged in a field experiment to explore the importance of managerial involvement in developing performance measurement narrowly understood to more active and analytical performance management.  Los Angeles is currently involved in a citywide reform intended to implement PerformanceStat-type management reforms in all 35 City Departments.  Recently, the City has embarked on a training program to introduce teams of managers to continuous learning techniques based on their organization’s PerformanceStat metrics.  In the field experiment groups of managers will be randomly assigned to control and experimental groups, where the experimental group will receive the continuous learning exercises.  The training will occur over the summer of 2016, with a follow-up survey scheduled for early fall 2016.  The follow-up survey follows upon a baseline administered to all field managers in Spring 2015.  The analysis will examine the extent to which training influenced managerial attitudes toward and use of data-driven management routines, as well as the extent to which these changes are mediated by managerial characteristics, such as public service motivation, measured at pre-test.