“Show me the Numbers”: Examining the Dynamics Between Evaluation and Government Performance in Developing Countries LESLI HOEY * University of Michigan, United States Summary. — This paper examines the dynamics between monitoring and evaluation (M&E) and government performance in developing countries, where M&E systems are expanding rapidly. Findings in Bolivia suggest that approaches to M&E can lower staff morale, create burdensome paperwork, blind managers to operational problems and emerging innovations, and reinforce self-censorship, contributing to the very problem M&E is intended to solve. Crafted appropriately, M&E can instead become a tool to build practical judgment, in- crease staff motivation, and improve implementation incrementally. Ultimately, these findings contribute to efforts to design M&E that can support staff working under complex working conditions. Ó 2014 Elsevier Ltd. All rights reserved. Key words — monitoring and evaluation, policy implementation, development planning, malnutrition, health policy, Bolivia 1. INTRODUCTION This paper examines the dynamics between monitoring and evaluation (M&E) and government performance in developing countries, where M&E systems are expanding rapidly (Ernesto, Shand, Mackay, Rojas, & Saaverdra, 2006; EvalPartners, 2014). Rather than asking about the technical quality and rigor of M&E being used (Fukuda-Parr, Greenstein, & Stewart, 2013), my questions build on research about the conditions under which such systems improve or erode development policy implementation (Hood, 2012). Based on a case study of Bolivia’s Zero Malnutrition (ZM) program, I suggest that mid-level managers may cling to collecting information about externally defined, quantitative indicators of staff performance as a reaction to complex social change processes—as coping mechanisms that give the allusion of controlling implementation. In these situations, evaluation can obscure operational issues, create burdensome paperwork, blind managers to emerging innovations, and rein- force self-censorship, contributing to the very problem M&E is intended to solve. On the other hand, where managers use M&E in ways that help build practical judgment about how to improve implementation, they create an environment con- ducive to learning—building motivation and trust—and engage more diverse actors and types of knowledge in analyz- ing problems and negotiating solutions. Ultimately, these findings contribute to literature aiming to reconsider how to design M&E to support staff working under complex condi- tions (Rogers & Fraser, 2014). Efforts to institutionalize government-based monitoring and evaluation 1 (M&E) systems in developing countries have grown considerably over the past decade in response to the Millennium Development Goals (Savedoff, Levine, & Birdsall., 2006), Poverty Reduction Strategy Papers (Holvoet, Gildemyn, & Inberg, 2012), and the Paris and Accra Declarations (High Level Forum, 2008; OECD/DAC Organization for Economic Coordination and Development/ Development Assistance Committee, 2005). Each of these ini- tiatives calls for more “country-owned” development (Hyden, 2008) as well as monitoring and evaluation of donor investments and public policies (Thomas, 2010). In response, international institutions have launched numerous evaluation networks (IOEC, 2014) and initiatives to build development evaluation capacity (Mackay, 2006; Naidoo, 2013; Savedoff et al., 2006). There is increasing evidence of “country-led— rather than donor-driven—efforts to institutionalize M&E” (May, Shand, Mackay, Rojas, & Saavedra, 2006, p. xi; Imas & Rist, 2009), and the demand for evaluators is growing; as of 2012, there were 138 national associations of professional evaluators representing 110 countries (EvalPartners, 2014), up from only five in 1990 when associations existed only in North America, Europe, and Australia (Donald, 2006). At the heart of this exponential growth is a belief that M&E serves a variety of purposes: to hold actors accountable, iden- tify policy options proven to work, and to improve the effec- tiveness of interventions during implementation (Hood, 2012; IOEC, 2014). While researchers—most notably those with the Abdul Latif Jameel Poverty Action Lab (J-PAL) (Kremer and Glennerster, 2012)—are showing that rigorous (e.g., randomized control trial) evaluations can help identify effective international development strategies, barriers still exist to support wider adoption of evidence-based policy during the planning process (Dhaliwal and Tulloch, 2011). Moreover, there is no standardized approach for building country-level capacity to mainstream M&E during the imple- mentation phase (Goldberg and Bryant, 2012). One of the major debates is about whether certain evaluation models may contribute to the very problem they are intended to solve * This study was supported primarily through an Inter-American Foun- dation Grassroots Development Fellowship as well as Field Research Grants awarded through Cornell University’s Tinker Program in Latin American Studies, Einaudi Center for International Studies, and Interna- tional Studies in Planning. The study’s sponsors played no role in the design of the study, in the collection, analysis, and interpretation of the data, in the writing of the report, or in the decision to submit the paper for publication. I am grateful for the support, assistance, and participation of staff at national, departmental, and municipal levels of Bolivia’s Ministry of Health and Zero Malnutrition Program and stakeholders from Boliv- ia’s international nutrition policy community. I also thank John Forester, David Pelletier, and Mark Constas for their feedback on an early version of this report, as well as two anonymous referees for their valuable co- mments and suggestions. Final revision accepted: December 18, 2014. World Development Vol. 70, pp. 1–12, 2015 0305-750X/Ó 2014 Elsevier Ltd. All rights reserved. www.elsevier.com/locate/worlddev http://dx.doi.org/10.1016/j.worlddev.2014.12.019 1