Essays on Frontline workers’ motivation and performance under public programs: A multi-method study

Abstract

Frontline workers constitute a critical part of the implementation cadre within public policies, serving a significant role in the delivery of public goods and services. This dissertation examines frontline workers’ work environment, the role of incentives and of prosocial behavior in affecting task performance, using different methodological lenses. Frontline behaviour and decision making have been studied in the literature on Street Level Bureaucracy (Lipsky, 1980) with limited clarity on ‘what comprises work.’ Furthermore, most studies are concentrated in policy implementation in the developed world. In Study 1, we examine frontline workers’ behaviour in the context of the implementation of health, nutrition and sanitation policies in India. This study provides a multi-stakeholder perspective into the nature of frontline work. Using the grounded theory approach (Strauss & Corbin, 1997), our findings suggest that the definition of what comprises work differs considerably between the workers, supervisors, and policy makers. Additionally, without proper incentives and no formal designation, workers are expected to be driven by beneficiary interests to perform towards policy goals. In Study 2, we investigate whether the nature of compensation offered to frontline workers alters prosocial effort? Frontline workers perform tasks that are uniformly prosocial; however, the incentives offered, differ across frontline cadres. While studies list a host of contexts where financial incentives influence prosocial effort differently (Gneezy, Meier, & Rey-Biel, 2011), there is limited clarity on the impact of the nature of compensation. In this study, we draw on agency theory (Eisenhardt, 1989), and the literature on prosocial effort, to conduct a randomized lab-in-the-field inquiry (Bozeman & Scott, 1992) with Anganwadi workers (workers under the nutrition policy) performing a real-effort task. The performance was adjudged on quantity and quality parameters, and the effort invested. Results demonstrate that uncompensated workers invest the greatest amount of effort while compromising on task quality, and vice-versa for subjects receiving pay-for-performance compensation. In Study 3, we probe the extent to which public and private benefit influence prosocial effort. Combining the research on prosocial performance improvement (Meier, 2006) with prospect theory (Tversky & Kahneman, 1992), specifically the literature on limited attention (Karlan, Connell, & Mullainathan, 2016) and loss aversion (Tversky & Kahneman, 1991), we designed interventions to potentially influence the performance of volunteers. We present causal estimates from a field experiment undertaken in western India with 1395 rural sanitation program volunteers, who are expected to encourage household-level toilet construction. The participants received monthly voice reminders framed to convey different degrees of public and private benefit. Accounting for the contextual limitations, the results suggest that private benefit messages influenced better task performance than public benefit messages. Together, the studies raise questions regarding the design of incentives and mismatch between incentives and nature of tasks performed. This thesis attempts to offer insights that aim to inform the implementation channel to enable improved frontline worker performance under public programs.