The Behavioral Professional IMPROVING DECISION-MAKING AND PERFORMANCE IN THE PUBLIC SECTOR JOANA S. LOURENÇO, RENOS VAKIS, AND LAURA ZORATTO We would like to thank Zeina Afif, Benu Bidani, Oscar Calvo-Gonzalez, Ana Maria Munoz Boudet, and Tasmia Rahman for their valuable comments and suggestions on earlier versions of this paper. Contents 05 01. The behavioral professional in practice. Looking at the mirror. 12 02. Connecting the dots. The three contexts of behavioral professionals. 14 03. From behavioral insights to solutions. 22 04. An unfinished agenda. Moving towards the behavioral professional. 23 References 27 Annex 1 The three contexts of behavioral professionals in detail 33 Annex 2 Common behavioral pitfalls among policy professionals Over the past decade, governments, multilateral organizations and think tanks have been increasingly using behavioral science as an additional tool to understand and tackle complex policy challenges in several sectors. Yet despite this increase in the use of behavioral science for policy design, little attention has been given so far to those individuals responsible for de- signing and implementing public policies and programs: policy professionals. Civil servants, like every human being, and against their best intentions, experience biases in information processing and are influenced by their mindsets as well as institutional and group contexts in multiple ways. This affects their decision-making, performance and, consequently, program design, implementation, and public services delivery. With the public sector respon- sible for 16 percent of total global employment, 30 percent of wage employment, and 38 per- cent of formal sector wage employment, the stakes could not be higher.1 Civil servants, like every human being, There are myriad ways in which high level deci- and against their best intentions, sion makers all the way down to frontline staff (social welfare workers, teachers, nurses, doc- experience biases in information tors, etc.) can affect access, quality and/or effec- processing and are influenced by their tiveness of service delivery to beneficiaries. For mindsets as well as institutional and example, recent work by the Mind, Behavior, and group contexts in multiple ways. Development Unit (eMBeD) and the World Bank Health, Nutrition, and Population Global Practice in a large-scale national study in the Philippines revealed differences in beliefs among frontline health and nutrition workers regarding the root causes of stunting. It showed that those who agreed with views of child stunting being due to genetic and racial factors (vs. nutrition and prenatal care) were less likely to work in facilities that provided better maternal and child nutrition services. Similarly, looking at the current challenges around the COVID-19 vaccine roll-out, health work- ers’ own mindsets and beliefs on vaccine efficacy and safety are likely to affect their motiva- tion as well as communication to beneficiaries. This is particularly critical for vaccine take-up given recent findings from vaccine hesitancy surveys that suggest that individuals see health sector workers as the most credible and trustworthy source of information on vaccine safety and efficacy. This note aims to achieve three objectives. First, it highlights recent examples building on work done by the eMBeD team and the World Bank at large on how behavioral bottlenecks can hinder key development goals, from ensuring inclusive and equitable education for all (SDG4) to ensuring good health and well-being (SDG3), among others. Second, the note proposes a behavioral framework highlighting the individual, group and institutional contexts that affect policy professionals. Finally, it showcases the relevance of the behavioral approach to a broad range of areas – including public service design, corruption and accountability, service design, access and delivery, civil servants’ performance – by pinpointing common bottlenecks faced, and potential solutions to overcome them. 1. Worldwide Bureaucracy Indicators (WWBI); Apolitical (2019) The Behavioral Professional I 4 01. The behavioral professional in practice. Looking at the mirror. Policy professionals are not immune to behavioral biases and are influenced by their social environments and their mindsets. This includes not only policymakers and civil servants working on implementation (e.g. Ministry of Education staff, procurement agency staff), but also frontline staff interacting directly with beneficiaries (e.g. nurses, teachers, social workers) who play a critical role in service delivery. Remarkably, a meta-analysis from nearly 500 studies by Durlak and DuPre (2008) re- vealed that the impact of programs or projects that are carefully implemented can be two to three times higher than programs that are not carefully implemented or have serious implementation problems. Whether a program is well implemented depends on frontline staff skill proficiency, but, as highlighted by the meta-analysis, also on a number of behav- ioral aspects. This includes policy professionals’ mindsets, their perceptions about the need for, and potential benefits of a given program, confidence in their ability to deliver, and sense of self-efficacy (Abry et al, 2013; Durlak & DuPre, 2008). Can these insights apply to policy professionals in development settings and impact their work? We highlight recent examples from work by eMBeD and other teams at the World Bank, in the context of different SDGs goals, which showcases how policy professionals’ behavioral biases, social environments, and mindsets can affect the achievement of de- velopment outcomes. We also illustrate through some of these examples how adopting a behavioral lens can complement existing solutions to support SDGs goals by improving program design and implementation through tackling policy professionals’ behavioral bi- ases, leveraging social dimensions, or re-designing processes. The Behavioral Professional I 5 HOW THE SYSTEM IS DESIGNED AFFECTS SDGs 1 and 8 FRONTLINE WORKER Improving Repayment MOTIVATION of Microcredit Loans In Bangladesh, eMBeD and the Social Protection Global Practice applied small behaviorally-informed changes to improve service delivery, by re-designing an existing process of microcredit loans repayments. Changes included simplifying the payment receipt form to minimize social workers’ effort, as well as using a feedback mechanism to make late payment by beneficiaries of the rural microcre- dit program salient. Preliminary results from two small pilot experiments suggest that these chang- es led to time savings (totaling seven hours each month) and a decrease in late payments (which also translates into additional cost and time savings, due to the need for fewer follow-up visits by social workers). This work is part of a World Bank project to improve the transparency and efficiency of programs – rural microcredit and two major cash transfer programs – for vulnerable populations by moderniz- ing social workers’ service delivery. The Behavioral Professional I 6 HEALTH WORKER BELIEFS AFFECT EFFORT AND SDG 2 ULTIMATELY QUALITY OF Improving Effectiveness SERVICE PROVISION of Nutrition Policies eMBeD and the Health, Nutrition, and Population Global Practice recently carried out a large-scale national study on stunting (an important marker of child undernutrition) in the Philippines. A mixed-method approach was used to explore beliefs and attitudes about stunting among frontline health and nutrition workers, local public officials, and caregivers, as well as how this affected the practices adopted by frontliners when providing care services. This included a listing and grouping exercise of root causes of stunting, as well as in-depth interviews and focus groups. A large-scale quantitative survey among frontline workers in 10 regions was then carried out to understand the importance given to the different beliefs that emerged during the initial exploratory work. Findings revealed that most workers’ knowledge and beliefs were consistent with conventional practices (i.e. emphasis on nutrition and prenatal care). However, a minority of workers also agreed with views of child stunting being due to genetic and racial factors, alongside faith- and fate-based factors. More importantly, these workers were less likely to work in facilities that provided better maternal and child nutrition services, suggesting that child outcomes and health workers beliefs are highly correlated (Sen et al., 2020). The Behavioral Professional I 7 UNCONSCIOUS BIASES SDG 3 CAN REDUCE PROGRAM/ SERVICE TAKE-UP Reducing Maternal and Neo-natal Mortality Rates In Haiti, low rates of prenatal and postnatal care, and institutional births, contribute to high levels of maternal and neo-natal mortality rates (the highest in the Latin America and Caribbean region). Research by eMBeD and the Health, Nutrition, and Population Global Practice revealed that pregnant women experience bar- riers for seeking and receiving prenatal care, including some related to uncon- scious biases from health workers. There are several ways in which frontline staff can affect access, quality and/or effectiveness of service delivery to beneficiaries. Specifically, the research carried out showed that pregnant women fear being judged negatively by nurses, as they are asked many questions at registration, including some on their sexual habits; this stereotype threat can prevent them from seeking care. Likewise, they face barriers that discourage them from delivering at an institution, such as lack of trust in medical staff due to rumors about medical staff negligence circulating around the community. Notably, while in other low-income countries 70 percent of births take place in a health facility, less than 40 percent do in Haiti. In another study with frontline health staff, Banerjee et al. (2008) showed that while a monitoring and financial penalties program was highly successful in in- creasing attendance rates of nurses in Indian public health facilities in the first six months, the program eventually became ineffective. This was due to actions by the local health administration staff who was responsible for enforcement of pen- alties (i.e. staff allowed nurses to claim an increasing number of “excused” absenc- es, possibly as a way to deal with pressures from nurses while meeting obligations towards their superiors in terms of program implementation). The Behavioral Professional I 8 TEACHER BELIEFS AFFECT THEIR PRACTICES SDG 4 AND ULTIMATELY Ensuring Inclusive and STUDENTS’ LEARNING Equitable Education for All Since teachers’ instructional practices and unconscious preferences can hin- der students’ gains and efforts, eMBeD is working with teachers in various settings to understand their own behavior in the classroom, identify potential biases and propose ways to overcome these. For example, in North Macedo- nia, eMBeD, together with the Social Protection and Jobs and Education Glob- al Practices tested different interventions to generate more inclusive school environments and tackle teachers’ biases by promoting socio-emotional skills in the classroom. A first intervention focused on fostering grit and reducing stereotype threat. Im- plemented nationally across all schools teaching in Macedonian and Albanian lan- guage, it targeted 6th and 7th grade students and their teachers. Promising results were observed from the intervention that trained teachers and where teachers delivered socio-emotional skills lessons to the students. Roma students in particu- lar showed positive impacts on socio-emotional skills and on grade point averages equivalent to three weeks of school. A second intervention, adding a mentorship model where 7th and 8th grade students mentored 4th grade students in the great er Skopje area showed promising impacts among girls and increased awareness by teachers of the value of socio-emotional skills for learning. A third intervention (underway as of printing) is focusing on primary school teachers to more directly tackle teacher’s classroom practices that can lead to unintentional discrimination of minority students. In Turkish schools, eMBeD is using Virtual Reality (VR) to improve empathy and overall attitudes of teachers towards Syrian students through perspective-chang- ing and contact theory. eMBeD is testing the effectiveness of a VR interventions in promoting empathetic teachers’ attitudes and behaviors toward different social groups (particularly Syrian students under Temporary Protection). In a first phase, exploratory fieldwork was carried out to examine empathy drivers of teachers and identify relevant content and format for the VR intervention. This was done through qualitative research with main actors, including teachers and students (Turkish and Syrian), but also principals and students’ families. Phase 2, will com- prise an evaluation of the effectiveness of the VR intervention, in which contact and perspective-taking will be used to allow teachers to experience and interact with Syrian students (the outgroup), or even become one. VR content focuses on increasing teachers’ motivation to use belonging techniques and equipping them with hands-on knowledge for doing so. The Behavioral Professional I 9 DECISION MAKERS’ NORMS, RECOGNITION, SDG 16 AND THEIR EFFECT Building Effective and ON PERFORMANCE Accountable Institutions In Romania, a behavioral diagnostic by eMBeD and the Governance Global Prac- tice identified psychological barriers for managers and other civil servants that hinder performance management. In the Romanian Civil Service, it was observed that the performance evaluation system was not being implemented as originally intended and was failing to effectively promote individual civil servant motivation and performance. To understand beliefs, motivations, and behaviors regarding performance apprais- al systems, the project included open-ended and guided focus group discussions with civil servants as well as in-depth interviews and vignettes with managers in seven different public organizations in Romania.2 The analysis revealed several bottlenecks. For example, most staff receive high ratings in annual assessments - the norm due to a risk of litigation and overall culture. The diagnostic found that meaningful ratings were hardly sought among managers, who preferred to maintain the status quo. This adversely affected motivation, especially of better performing civil servants. Building on the evidence gathered and insights from behavioral sci- ence, eMBeD devised a set of actionable solutions that could be tested to improve motivation and engagement of employees. For instance, having mid-term check-in with coaching of (technical) managers (e.g. on giving feedback, using behaviorally informed motivational tools such as social recognition and appreciation). In Nigeria, although record keeping in health facilities is an important task (e.g. it allows the government to track cash flows), it was perceived as a mundane task by workers who lacked the motivation to complete their tasks. eMBeD test- ed the effectiveness of a social recognition intervention to improve record keeping by civil servant in health facilities in two Nigerian states. An inexpensive four-week social recognition intervention – weekly “Certificate of Excellence” showing performance through a number of “stars” displayed promi- nently within the facility, coupled with distinction for the best performing facility and all staff members in a special ceremony with the Permanent Secretary of Health at the end – increased performance in one of the two states (Gauri et al., 2019).3 2. Vignettes are short stories that present circumstances similar to what respondents might face and asks them about the choice that the main character in the story would make. This response is taken to be indicative of the choice that the respondents would make themselves. 3. Differences across states were not accounted by observable characteristics at the facility level, but closer examination of social contexts and institutional structures would be required to examine why findings didn’t generalize. The Behavioral Professional I 10 SUNK COST BIAS AND INERTIA AFFECT SDG 16 BUDGET PREPARATION Improving Resource Allocation The final example shows how policy professionals biases can affect program management and investment, or more specifically, how fund allocation deci- sions can be impacted by sunk cost bias (percentage of budget already spent). When development professionals were presented with a decision on whether to continue committing funds to an underperforming program that had been active for four years, all things being equal, they were significantly more likely to do so in a scenario where 70% ($350 million) of the funds had been spent vs. when only 30% ($150 million) of the funds had been spent (World Bank, 2015). This example shows that to reduce budget inertia – a typical public financial management prob- lem - sunk-cost bias of budget officials needs to be addressed. The Behavioral Professional I 11 02. Connecting the dots. The three contexts of behavioral professionals. Policy professionals do not behave in vacuum; their decision-making process is influenced by three interdependent contexts: institutional, individual and group (Figure 1).4 The institutional context refers to how systems are set in each setting and organization. It pertains specifically to how rules and processes are designed, implemented, and monitored. The group context refers to the role of social dimensions, such as the individual’s social iden- tity, group identity, and social norms or rules and standards regarding behaviors in a society. The individual context encompasses dimensions pertaining to one’s motivation to perform, interpretation of information (data, others or situations) and abilities. It covers multiple as- pects, such as the role of monetary and non-monetary rewards, beliefs about how the world works, cognitive biases, or non-cognitive or soft skills, among others. These three contexts are interdependent. For instance, beliefs, motivation, and social iden- tities can influence behavior towards beneficiaries. Likewise, organizational culture and bu- reaucratic norms can help shape decisions and actions of civil servants. Within each of these contexts, several factors are at play, as detailed in Annex 1. 4. The proposed framework builds on literature on behavioral science, public administration and service delivery, as well as on existing frameworks relevant for the context at hand. These include the Mind, Society, and Behavior Framework on how individuals make decisions (i.e. people think automatically, think socially, and think with mental models) (World Bank, 2015) and the Means, Motives and Opportunity Framework on community health worker performance (John et al, 2019). The Behavioral Professional I 12 FIGURE 1: THE BEHAVIORAL POLICY PROFESSIONAL’S CONTEXTS • “System” Design: • Social Identity TEXT Processes and Rules N • Empirical Expectations • Performance Assessment: L CO • Normative Expectations Monitoring and A ON Feedback Cycles I UT • Enabling TIT Environment Effort INS and Culture Motivation GROU Perseverance P CON Decision accuracy Performance TEXT Absenteeism, Etc. XT TE N O C AL IDU INDIV ONE’S MOTIVATION ONE’S INTERPRETATION ONE’S ABILITY • Goals and Career Aspirations • Mental Models • Knowledge and Beliefs • Monetary and • Non-cognitive skills Non-Monetary Incentives • Cognitive Biases • Technical skills • (Perceived) Relevance • Bandwidth of Work • Civic-Mindedness The Behavioral Professional I 13 03. From behavioral insights to solutions. The Behavioral Professional I 14 RETHINKING POLICY DESIGN BEHAVIORALLY 42 Solutions Table 1 below presents a theory of change of sort, summarizing selected ex- amples of relevant policy areas and moments of the policy cycle that can, and typically are, influenced by biases (see also Annex 2 for a description of typical biases among policy professionals) and other behavioral barriers. It then outlines some possible solutions. The examples come from a range of poli- cy areas including budget and investment planning, corruption and accountability, policy and program design, access to services, and civil servants’ performance. Potential solutions – at the individual, group and institutional levels5 – are also presented that have been shown to influence similar behaviors in other contexts and that could be applied to improve the decision-making, implementation and service delivery of policies and programs. 5. To illustrate, systems can for example be modified using principles of choice architecture, to mitigate errors of reasoning and biases that can have direct impacts on program outcomes (addressing the institutional context); social recognition and appreciation can be used to increase motivation and bureaucratic performance (addressing the individual context); role models or interaction with social referents, as well as identity-labelling interventions, can be used to change norms (addressing the group context). The Behavioral Professional I 15 TABLE 1. (1/4) COMMON BARRIERS AFFECTING THE DECISION-MAKING PROCESS AND POTENTIAL SOLUTIONS COMMON BIASES / POTENTIAL SOLUTIONS TO AFFECTING… BEHAVIORAL ACROSS THE THREE CONTEXTS… IMPROVE… BARRIERS Encourage cognitive strategies to evaluate evidence (e.g. “consider the opposite,” “taking an outsider view”) How policy Confirmation bias, proposals and Prime individuals to “think critically” overconfidence, programs are “better than the discussed; how Use debiasing checklists and prompt average” effect, evidence is their application Policy and blind spot presented; how program design bias, illusion of opposing or Leverage technology to debias using active and similarity, group minority views are personalized learning and repeated practice reinforcement expressed and considered Gather viewpoints anonymously in advance to enable divergent views (e.g. anonymous voting) Have leaders in the organization publicly promote divergent thought Communicate descriptive norms to correct misperceptions, or conveyed positive dynamic trends (reduction in corruption level) Increase or make the moral costs more salient (e.g. by decreasing the ambiguity in evaluation standards) Increase accountability to beneficiaries and observability (e.g. reduce frictions for access to information and exercising voice) Misperceived Factors that may level of corruption contribute to Promote accountability and transparency to among peers, corruption in the increase trust (e.g. using community leaders Allocation of reciprocity, social form of bribery or to aid in targeting, increasing flow of information resources and preferences, other rent-seeking to beneficiaries) public goods system of behavior that could Create anonymous reporting mechanisms; en- provision incentives, moral undermine policy sure security of staff reporting corruption costs implementation and rent-seeking behavior Increase attention to standards for honesty through reminders (e.g. sign moral/accuracy- confirming statement upfront) Create conflict of interest forms to be filled and confirmed regularly Exposure to positive role models The Behavioral Professional I 16 TABLE 1. (2/4) COMMON BARRIERS AFFECTING THE DECISION-MAKING PROCESS AND POTENTIAL SOLUTIONS COMMON BIASES / POTENTIAL SOLUTIONS TO AFFECTING… BEHAVIORAL ACROSS THE THREE CONTEXTS… IMPROVE… BARRIERS Encourage employee intrinsic motivation (e.g. gratitude expressions, publicly acknowledging good work behavior; organizational awards for outstanding performance) Decrease costs for providing feedback Barriers to effective (e.g., timely prompts, defaults, clear and performance Formal and infor- transparent rules) management mal rules, social and procedures Provide support to actions by managers (e.g. Performance norms, moral which can coaching on giving feedback; “cheat sheet” management licensing, mental impact employee with tips and tools to engage and motivate models, trust motivation and staff); mentorship programs engagement Create department and individual-level goals and promote a results-oriented mindset among staff (away from a focus on process) Recognize improvements in performance towards a goal/leverage goal gradient effect ‘Premortems’ to imagine project failure and root causes Encourage looking to similar project outside of own context or taking an outsider view to How project increase objectivity Optimism bias, decisions and Program planning fallacy, implementation Prompt estimates according to different and project sunk cost bias, can be objectively scenarios to encourages thinking about a management normalcy bias, monitored; larger number of possibilities and investment status quo bias openness to adapt planning to new processes Build in breakpoints to revisit assumptions and plans Establish mechanisms for (more frequent) feedback The Behavioral Professional I 17 TABLE 1. (3/4) COMMON BARRIERS AFFECTING THE DECISION-MAKING PROCESS AND POTENTIAL SOLUTIONS COMMON BIASES / POTENTIAL SOLUTIONS TO AFFECTING… BEHAVIORAL ACROSS THE THREE CONTEXTS… IMPROVE… BARRIERS Leaders publicly promote value of service providers work Soft skills training (e.g. goal setting, grit, belonging) to improve commitment, and public service delivery How frontline Preferences staff understand Use non-monetary incentives (e.g. social and beliefs, and believe in the recognition, gratitude expression) to improve Service delivery perceived value objective of the motivation/performance quality and of the work, social work; perceived fidelity of program norms, incentives, need for service; Recognize improvements in performance implementation accountability motivation and job towards a goal satisfaction Increase accountability to beneficiaries and observability (e.g. improve access information and exercising voice) Use social comparison nudges to prompt individuals to complete specific tasks or adhere to a target behavior Leverage technology to debias using active and personalized learning and repeated practice Communicate descriptive norms to correct misperceptions Promote accountability and transparency (e.g. using community leaders to aid in targeting, Preferences and increasing flow of information to beneficiaries) beliefs, racial The quality and ethnic biases of interaction Encourage interactions with other service providers that could be leveraged as role Service delivery and stereotypes between social models quality and access related to workers and out-group beneficiaries Implement strategies which facilitate and beneficiaries improve objective assessment (e.g. blinding of gender or ethnicity of applicants from case files) Use role models to signal an appropriate behavior one is trying to promote Encourage perspective-taking The Behavioral Professional I 18 TABLE 1. (4/4) COMMON BARRIERS AFFECTING THE DECISION-MAKING PROCESS AND POTENTIAL SOLUTIONS COMMON BIASES / POTENTIAL SOLUTIONS TO AFFECTING… BEHAVIORAL ACROSS THE THREE CONTEXTS… IMPROVE… BARRIERS Reduce cost of learning the actions to follow How well one to deliver the program (e.g. simplify information, can understand make information more direct, salient, visual, and interpret etc.) information about Reduce friction costs to adhere to steps in the program rules Information costs, program implementation or beneficiary Service delivery mental bandwidth, eligibility; how quality and limited attention, compliance with Simplify forms and procedures or use checklists fidelity of program perceived cost, instructions/rules implementation self-efficacy is perceived (easy, difficult); existence Remove unnecessary information updates, of unnecessary documentation, calculations steps or actions required Send timely reminders to prompt action The Behavioral Professional I 19 (RE)TRAINING BEHAVIORALLY Addressing Cognitive and Motivational Competencies Potential entry points for capacity building training and service delivery improve- ments include debiasing, with focus on biased beliefs regarding context or bene- ficiaries or biases affecting data interpretation. The evidence generally suggests that informing people of biases is not enough. Individuals still need to be able to recognize situations where biases may be at play as well as be motivated and able to effectively use relevant strategies (Chang et al, 2016; Neilens et al, 2009; Soll et al, 2014; Hallsworth et al, 2018). Training can focus on teaching people appropriate rules and principles and changing the individual’s cognitive strategies (Soll et al, 2014; Larrick, 2004). This holds similarities to the concept of “boosts”, which aim to foster the individual’s cognitive and/or motivational competences to promote be- havioral change and require some level of engagement (Hertwig & Grüne-Yanoff, 2017). For instance, a recent survey experiment with policy professionals showed that deliberation in pairs can reduce confirmation bias (Banuri et al, 2019; see also Vivalt and Coville, 2019). What type of training format could help policy professionals? Recent evidence suggests that while unconscious (or implicit) bias training is effective at increasing awareness, its impact in behavior in the workplace, relating to equality, diversity and inclusion, is limited (Alleyne, 2020; Atewologun, Cornish & Tresh, 2018). In line with this, a recent randomized evaluation testing the effectiveness of a one- off online diversity training in a global organization found some positive effects on attitudes, but mostly null effects on employee behavioral change (Chang et al., 2019). There is, however, suggestive evidence that aspects such as targeted and repeated practice, personalized and timely feedback, and domain-specific train- ing may increase efficacy of debiasing training (Chang et al, 2016; Soll et al, 2014). Importantly, technology offers several possibilities to implement these strategies. First, recent studies showed that a one-shot debiasing training intervention, imple- mented through a “serious game” making use of repetition, personalization and feedback led to positive effects both immediately and up to three months from training (Morewedge et al, 2015; Sellier et al, 2019). Second, VR technology can be used for immersive training, while simulating real life conditions and affording opportunities for repeated practice. Recent evidence suggests that VR training holds promise as a mechanism to develop technical, practical, and socio-emotional skills in students (Angel-Urdinola, Castillo-Castro, & Hoyos, 2021). Also, as mentioned above, eMBeD is testing the effectiveness of a VR interventions in promoting empathetic teachers’ attitudes and behaviors The Behavioral Professional I 20 towards Syrian students in Turkey. Third, mobile technology can be leveraged in combination with behavioral strate- gies to build personalized, sustainable, and scalable behavioral change solutions, while allowing to monitor user’s responses and (sustained) engagement.6 Within a technological able environment, multiple behavioral science tools can be applied either to, for example, motivate users towards a goal, including personalized feed- back, reminders for action, actionable tips to make the intended behavior simpler or remove triggers, tools to build resilience such as work around self-efficacy, per- ceived control or growth mindset, timely and positive reinforcement, peer support and accountability. In line with the available evidence discussed above, careful design – including by building both on behavioral science and learning principles – and testing of training interventions is essential. 6. An area that has seen a considerable number of smartphone application for behavioral change is weight loss, and similar strategies can be tested to promote behavioral change in other areas. For instance, recent research shows promising results for Noom, a commercial app that offers daily informational reminders, tracking, virtual 1:1 behavior change coach, and support group, using psychological principles to encourage behavioral change (eChin et al. 2016; DeLuca et al., 2020). The Behavioral Professional I 21 04. An unfinished agenda. Moving towards the behavioral professional. Lessons drawn from behavioral science are increasingly being used to better under- stand and tackle intractable policy challenges. Nonetheless, dedicated, knowledgeable, well-meaning policy professionals are also subject to beliefs and attitudes that can result in biases, shaping their actions and impacting choices and success of policies or programs. This note highlights how behavioral science can be used to mitigate such biases in policy and implementation contexts, but also for addressing bottlenecks linked with employees’ motivations and organizational systems and processes that can shape policy profession- als’ attitudes and behaviors. Providing policymakers and program implementers with more tools for effective program design and implementation, and with tools on how to use them and apply to their views of problems and solutions, can be an effective addition to existing strategies and plans and it may be the key to unlocking developing impacts at scale around the world. The Behavioral Professional I 22 References A health care system. Journal of the European Economic Associ- ation, 6(2-3), 487-500. Abry, T., Rimm-Kaufman, S. E., Larsen, R. A., & Brewer, A. J. (2013). The influence of fidelity of implementation on teacher– Banuri, S., & Keefer, P. (2015). Was Weber right? The effects of student interaction quality in the context of a randomized con- pay for ability and pay for performance on pro-social motivation, trolled trial of the Responsive Classroom approach. Journal of ability and effort in the public sector. The World Bank. School Psychology, 51(4), 437-453. Banuri, S., Dercon, S., & Gauri, V. (2019). Biased policy profes- Alleyne, A. “Our obsession with unconscious bias created a di- sionals. The World Bank Economic Review, 33(2), 310-327. versity disaster”. Wired. November 11, 2020. https://www.wired. Battaglio Jr, R. P., Belardinelli, P., Bellé, N., & Cantarelli, P. (2019). co.uk/article/unconscious-bias-diversity-fix Behavioral public administration ad fontes: A synthesis of re- Andersen, S. C., & Guul, T. S. (2019). Reducing minority discrim- search on bounded rationality, cognitive biases, and nudging in ination at the front line—combined survey and field experimen- public organizations. Public Administration Review, 79(3), 304- tal evidence. Journal of Public Administration Research and 320. Theory, 29(3), 429-444. Bellé, N., Cantarelli, P., & Belardinelli, P. (2017). Cognitive biases Angel-Urdinola, D. F., Castillo-Castro, C., & Hoyos, A. (2021). in performance appraisal: Experimental evidence on anchoring Meta-Analysis Assessing the Effects of Virtual Reality Training and halo effects with public sector managers and employees. on Student Learning and Skills Development. Draft Working Pa- Review of Public Personnel Administration, 37(3), 275-294. per. World Bank. Bellé, N., Cantarelli, P., & Belardinelli, P. (2018). Prospect theo- Apolitical (2019). From gender gaps to corruption: 5 lessons ry goes public: Experimental evidence on cognitive biases in from the first public work survey. Retrieved from https://apo- public policy and management decisions. Public Administration litical.co/en/solution_article/from-gender-gaps-to-corruption-5- Review, 78(6), 828-840. lessons-from-the-first-government-work-survey Bergman, P., Lasky-Fink, J., & Rogers, T. (2019). Simplification Ashraf, N., Bandiera, O., & Jack, B. K. (2014). No margin, no mis- and defaults affect adoption and impact of technology, but de- sion? A field experiment on incentives for public service deliv- cision makers do not realize it. Organizational Behavior and Hu- ery. Journal of public economics, 120, 1-17. man Decision Processes. Ashraf, N., Bandiera, O., & Lee, S. S. (2015). Do-gooders and Bhanot, S. P., & Linos, E. (2020). Behavioral Public Administra- go-getters: career incentives, selection, and performance in tion: Past, Present, and Future. Public Administration Review, public service delivery. Harvard Business School. 80(1), 168-171. Atewologun, D., Cornish, T., & Tresh, F. (2018). Unconscious Bicchieri, C., & Xiao, E. (2009). Do the right thing: but only if bias training: An assessment of the evidence for effectiveness. others do so. Journal of Behavioral Decision Making, 22(2), 191- Equality and Human Rights Commission Research Report Se- 208. ries. Bold, T., Kimenyi, M., Mwabu, G., & Sandefur, J. (2018). Exper- B imental evidence on scaling up education reforms in Kenya. Journal of Public Economics, 168, 1-20. Baekgaard, M., Christensen, J., Dahlmann, C. M., Mathiasen, Bteddini, Lida; Hasnain, Zahid; Kay, Kerenssa Mayo; Rogger, A., & Petersen, N. B. G. (2019). The role of evidence in politics: Daniel Oliver. 2018. Civil Servant Surveys help assess Govern- Motivated reasoning and persuasion among politicians. British ment Capability in a Program for Results (English). Governance Journal of Political Science, 49(3), 1117-1140. Notes; no. 10. Washington, D.C. : World Bank Baker, S. G., Patel, N., Von Gunten, C., Valentine, K. D., & Scher- Buell, R. W., Kim, T., & Tsay, C. J. (2017). Creating reciprocal er, L. D. (2020). Interpreting politically-charged numerical infor- value through operational transparency. Management Science, mation: The influence of numeracy and problem difficulty on 63(6), 1673-1695. response accuracy. Judgment & Decision Making, 15(2). Butler, D. M., & Broockman, D. E. (2011). Do politicians racially Banerjee, A. V., Duflo, E., & Glennerster, R. (2008). Putting a discriminate against constituents? A field experiment on state band-aid on a corpse: incentives for nurses in the Indian public The Behavioral Professional I 23 legislators. American Journal of Political Science, 55(3), 463- American journal of community psychology, 41(3-4) 477. E C Edmondson, A. C. (2018). The fearless organization: Creating Cable, D. M., Gino, F., & Staats, B. R. (2013). Breaking them psychological safety in the workplace for learning, innovation, in or eliciting their best? Reframing socialization around new- and growth. John Wiley & Sons. comers’ authentic self-expression. Administrative science quarterly, 58(1), 1-36. G Chang, W., Chen, E., Mellers, B., & Tetlock, P. (2016). Devel- Gauri, V., Jamison, J. C., Mazar, N., & Ozier, O. (2019). Moti- oping expert political judgment: The impact of training and vating bureaucrats through social recognition: External valid- practice on judgmental accuracy in geopolitical forecasting ity—A tale of two states. Organizational Behavior and Human tournaments. Judgment and Decision making, 11(5 Decision Processes. Chang, E. H., Milkman, K. L., Gromet, D. M., Rebele, R. W., Gneezy, U., Saccardo, S., & Van Veldhuizen, R. (2019). Bribery: Massey, C., Duckworth, A. L., & Grant, A. M. (2019). The mixed Behavioral drivers of distorted decisions. Journal of the Euro- effects of online diversity training. Proceedings of the Nation- pean Economic Association, 17, 917-946. al Academy of Sciences, 116(16), 7778-7783. Grant, A. M. (2008). Does intrinsic motivation fuel the proso- Coyle-Shapiro, J. A. M. (2002). A psychological contract per- cial fire? Motivational synergy in predicting persistence, per- spective on organizational citizenship behavior. Journal of Or- formance, and productivity. Journal of applied psychology, ganizational Behavior: The International Journal of Industrial, 93(1), 48. Occupational and Organizational Psychology and Behavior, 23(8), 927-946. H Hallsworth, M., Egan, M., Rutter, J. & McCrae, J. (2018). Using Croskerry, P., Singhal, G., & Mamede, S. (2013). Cognitive debi- behavioural science to improve how governments make deci- asing 1: origins of bias and theory of debiasing. BMJ Qual Saf, sions. UK Behavioral Insights Team. 22(Suppl 2), ii58-ii64. Hasnain, Zahid; Rogger, Daniel Oliver; Walker, Daniel John; D Kay, Kerenssa Mayo; Shi, Rong. 2019. Innovating Bureaucracy Delgado, L., & Shealy, T. (2018). Opportunities for greater en- for a More Capable Government (English). Washington, D.C.: ergy efficiency in government facilities by aligning decision World Bank Group. structures with advances in behavioral science. Renewable Hauser, O. P., Linos, E., & Rogers, T. (2017). Innovation with and Sustainable Energy Reviews, 82, 3952-3961. field experiments: Studying organizational behaviors in actual Devarajan, S. (2014). What the 2004 WDR Got Wrong. The organizations. Research in Organizational Behavior, 37, 185- World Bank – Future of Development, viewed November 198. 25, https://web.worldbank.org/archive/website01605/WEB/ Hertwig, R., & Grüne-Yanoff, T. (2017). Nudging and boosting: WHAT_200.HTM Steering or empowering good decisions. Perspectives on Psy- DiMaggio, P., & Markus, H. R. (2010). Culture and social psy- chological Science, 12, 973-986. chology: Converging perspectives. Social Psychology Quar- Hollingworth & Barker (2020). How to Debias Your Organisa- terly, 73(4), 347-352. tion. The Behavioral Architects. Dudley, S. E., & Xie, Z. (2020). Designing a Choice Architecture for Regulators. Public Administration Review, 80(1), 151- 156. J Jackson, D., & Köbis, N. (2018). Anti-corruption through a so- Durlak, J. A., & DuPre, E. P. (2008). Implementation matters: cial norms lens. U4 Issue, 2018(7). A review of research on the influence of implementation on program outcomes and the factors affecting implementation. The Behavioral Professional I 24 John, A., Newton-Lewis, T., & Srinivasan, S. (2019). Means, Psychological Science, 23(7), 710-717. Motives and Opportunity: determinants of community health Montibeller, G., & Von Winterfeldt, D. (2015). Cognitive and worker performance. BMJ global health, 4(5), e001790. motivational biases in decision and risk analysis. Risk analysis, 35(7), 1230-1251. Montibeller, G., & Von Winterfeldt, D. (2015). K Cognitive and motivational biases in Kahneman, D., Rosenfield, A. M., Gandhi, L., & Blaser, T. (2016). Noise: How to Overcome the High, Hidden Cost of Inconsis- Morewedge, C. K., Yoon, H., Scopelliti, I., Symborski, C. W., tent Decision Makinghttps. hbr. org/2016/10/noise. Korris, J. H., & Kassam, K. S. (2015). Debiasing decisions: Im- proved decision making with a single training intervention. Kalaj, J., Rogger, D., & Somani, R. (2020). Bureaucrat Time-Use Policy Insights from the Behavioral and Brain and Productivity: Evidence from a Survey Experiment. Kautz, T., Heckman, J. J., Diris, R., Ter Weel, B., & Borghans, N Nas Ozen, E., Hut, S., Levin, V., & Munoz Boudet, A. M. (2020). L. (2014). Fostering and measuring skills: Improving cogni- A Field Experiment on the Role of Socioemotional Skills and tive and non-cognitive skills to promote lifetime success (No. Gender for Hiring in Turkey. w20749). National Bureau of Economic Research. Nathan, N. L., & White, A. (2019). Experiments on and with Kay, K., Rogger, D., & Sen, I. (2020). Bureaucratic Locus of Street-Level Bureaucrats. Control. Neggers, Y. (2018). Enfranchising your own? experimental Keulemans, S., & Van de Walle, S. (2020). Understanding evidence on bureaucrat diversity and election bias in India. street-level bureaucrats’ attitude towards clients: Towards a American Economic Review, 108(6), 1288-1321. measurement instrument. Public Policy and Administration, 35(1), 84-113. Neilens, H. L., Handley, S. J., & Newstead, S. E. (2009). Effects of training and instruction on analytic and belief-based rea- König-Kersting, C., Pollmann, M., Potters, J., & Trautmann, S. T. soning processes. Thinking & Reasoning, 15(1), 37-68. (2017). Good decision vs. good results: Outcome bias in finan- cial agents’ rewards. NITI Aayog (2018). Aspirational districts: unlocking potentials. https://niti.gov.in/sites/default/files/2018-12/AspirationalDis- L tricts-Book.pdf Larrick, R. P. (2004). Debiasing. Blackwell handbook of judg- Nørgaard, A. S. (2018). Human behavior inside and outside ment and decision making, 316-338. bureaucracy: Lessons from psychology. Journal of Behavioral Liaqat, A. (2019). No Representation without Information: Poli- Public Administration, 1(1). tician Responsiveness to Citizen Preferences. O Lipsky, M. (2010). Street-level bureaucracy: Dilemmas of the Okeke, E. N., & Godlonton, S. (2014). Doing wrong to do right? individual in public service. Russell Sage Foundation. Social preferences and dishonest behavior. Journal of Eco- Liu, X., Stoutenborough, J., & Vedlitz, A. (2017). Bureaucrat- nomic Behavior & Organization, 106, 124-139. ic expertise, overconfidence, and policy choice. Governance, 30(4), 705-725. P Paarlberg, L. E., Perry, J. L., & Hondeghem, A. (2008). From M theory to practice: Strategies for applying public service mo- Mangla, A. (2015). Bureaucratic norms and state capacity in In- tivation. Motivation in public management: The call of public dia: implementing primary education in the Himalayan region. service, 268-293. Asian Survey, 55(5), 882-908. Pepinsky, T. B., Pierskalla, J. H., & Sacks, A. (2017). Bureaucra- Milkman, K. L., Akinola, M., & Chugh, D. (2012). Temporal dis- cy and service delivery. Annual Review of Political Science, 20, tance and discrimination: An audit study in academia. 249-268. The Behavioral Professional I 25 R V Relationality, Equality of Agency, and Development, qin V. Cul- Vivalt, E., & Coville, A. (2019). How do policymakers update?. ture and Public Action, The International Bank for Reconstruc- Unpublished manuscript, Berkeley, CA: University of Califor- tion and Development, The World Bank, Washington, DC. nia, Berkeley. Reinikka, R., & Svensson, J. (2001). Explaining leakage of pub- W lic funds. The World Bank. Westphal, J. D., & Zajac, E. J. (2013). A behavioral theory of Ritz, A., Brewer, G. A., & Neumann, O. (2016). Public service corporate governance: Explicating the mechanisms of socially motivation: A systematic literature review and outlook. Public situated and socially constituted agency. Academy of Man- Administration Review, 76(3), 414-426. agement Annals, 7(1), 607-661. Rogger, D., & Somani, R. (2019). Hierarchy and information. World Bank. 2015. World Development Report 2015: Mind, So- The World Bank. ciety, and Behavior. Washington, DC: World Bank World Bank. 2015. World Development Report 2017: Gover- S nance and the Law. Washington, DC: World Bank Sanna, L. J., Schwarz, N., & Small, E. M. (2002). Accessibility experiences and the hindsight bias: I knew it all along versus World Bank. 2018. Even Teachers are Bested by Their Bi- it could never have happened. Memory & Cognition, 30(8), ases (English). eMBeD brief. Washington, D.C. : World 1288-1296. Bank Group. http://documents.worldbank.org/curated/ en/976931525330213869/Even-Teachers-are-Bested-by- Scharbatke-Church, C., & Chigas, D. V. (2019). 11 Using sys- Their-Biases tems thinking to understand and address corruption in the criminal justice system in fragile states. Corruption, Social World Bank. 2018b. Motivating Public Sector Workers in Sciences and the Law: Exploration across the disciplines, 176. Nigeria (English). eMBeD brief. Washington, D.C. : World Bank Group. http://documents.worldbank.org/curated/ Sellier, A. L., Scopelliti, I., & Morewedge, C. K. (2019). Debias- en/440211517949747866/Motivating-Public-Sector-Work- ing training improves decision making in the field. Psycholog- ers-in-Nigeria ical science, 30(9), 1371-1379. World Bank. Upcoming. BRIEF: Social Workers for Local Ser- Sen, I., Mbuya, N., Demombynes, G., & Gauri, V. (2020). Mind vice Delivery in Social Protection. Over Matter in the Philippines: A Study of Key Stakeholders’ Perceptions of Childhood Stunting. Wright, B. E., & Davis, B. S. (2003). Job satisfaction in the pub- lic sector: The role of the work environment. The American Slough, T. (2018). Bureaucrats Driving Inequality in Access: review of public administration, 33(1), 70-90. Experimental Evidence from Colombia. http://taraslough.com/ assets/pdf/JMP.pdf Z Soll, J. B., Milkman, K. L., & Payne, J. W. (2014). A user’s guide Zahid Hasnain and Daniel Rogger (2018) “Innovating Bureau- to debiasing. cracy for Increasing Government Productivity” World Bank Governance Note. Washington, D.C. : World Bank Group Stanovich, K. E., & West, R. F. (2008). On the relative inde- pendence of thinking biases and cognitive ability. Journal of Zhao, H. A. O., Wayne, S. J., Glibkowski, B. C., & Bravo, J. personality and social psychology, 94(4), 672. (2007). The impact of psychological contract breach on work-related outcomes: a meta-analysis. Personnel psychol- Symborski, C., Barton, M., Quinn, M. M., Korris, J. H., Kassam, ogy, 60(3), 647680. K. S., & Morewedge, C. K. (2017). The design and development of serious games using iterative evaluation. Games and Cul- ture, 12(3), 252-268. The Behavioral Professional I 26 Annex 1 The three contexts of behavioral professionals in detail supervisors and field staff, respectively), such as INSTITUTIONAL CONTEXT principal’s (e.g. supervisors) inability to effectively monitor performance and agent’s (e.g. field staff) poor compliance with implementation or avoid- How the “System” is Designed: ance of formal duties (Nathan & White, 2019; Ba- nerjee et al, 2008). Processes and Rules Accountability to beneficiaries and observability of The way institutions are set and organized effort will also affect performance. Findings from (rules, procedures, systems for information Reinikka and Svensson (2001) nicely illustrate this flow) affects directly decision-making by influ- by showing that leakage of public funds for ed- encing how policy professionals understand, ucation at the local government level in Uganda adhere and follow specific parameters set. A was dramatically reduced by decreasing the cost clear, transparent and monitored process can lead of acquiring information and exercising voice. to different decisions and behaviors than vague, At the level of performance assessments, biases ad hoc one. This can affect anything from strategy can result in incorrect evaluations of civil servants design, delegating tasks, hiring, to performance that may negatively affect performance. For exam- tracking (e.g. Delgado & Shealy, 2018). ple, experimental studies completed with public servants have documented the presence of halo effects, whereby those who were perceived to How Performance have higher skills along one dimension, were giv- en higher scores on other dimensions, and there- is Assessed: fore overall performance (Bellé et al. 2017, 2018). Performance Monitoring Outcome bias – or evaluation of the quality of de- cisions based on their outcomes with little regard and Feedback Cycles for the presence of uncertainty at the time the de- cision was made – may also introduce distortions Incentives (see individual context below) and (König-Kersting et al, 2017). monitoring ability affect program administra- tion and implementation by policy professionals. Finally, policy professionals are challenged in Policies or programs may fail or deviate from their terms of receiving asymmetric feedback. Namely, objectives due to principal-agent problems (e.g. missing, infrequent, or delayed feedback regard- The Behavioral Professional I 27 ing the impact of decisions is likely to limit the abil- ity to correct bias and improve decisions (Dudley INDIVIDUAL CONTEXT & Xie, 2020). One’s Motivation Enabling Environment This relates to how factors linked with the indi- vidual’s desire to perform – such as goals and and Culture aspirations, expectations regarding the employ- ment relationship, monetary and non-monetary Institutional culture and the hierarchy of social rewards, perceptions about the need for the roles can also affect policy professionals’ atti- program/service, among others – affect his/her tudes and behaviors (Galley et al. 2013). There decision-making and behavior. are several other aspects of the work environ- ment that managers can maneuver to shape em- ployee performance. These include transparen- GOALS AND CAREER ASPIRATIONS cy and openness, which can promote increased perceptions of fairness and encourages feedback Goal setting as part of the performance evaluation through more regular and honest communication process has been shown to be important in the between manager and civil servants. These con- public sector in that it can offer a tool for manag- ditions are referred to as “psychological safety,” ers to motivate and improve performance (Paarl- where team members hold beliefs of mutual trust berg et al, 2008). and respect for each other, are comfortable being themselves, and believe that they will not be pun- Promotion prospects and career advancement ished unduly for speaking their mind (Edmondson, can also matter for performance. For instance, in 2018). Psychological safety is an important predic- a study in Zambia, Ashraf et al. (2015) showed that tor of performance. career incentives at the recruitment stage affect- ed who self-selected into a public health job (skills Organizational incentives, such as management and ambitions), attracting health workers who practices rewarding information acquisition, have completed 29% more household visits and posi- also been shown to mediate how policy officials tively impacting program outcomes. receive and use new information to reduce errors in the information used to make decisions (Rogger Another important dimension relates to the ex- & Somani, 2019). Additionally, institutions are so- tent to which employers meet the psychological cial structures and norms can affect behavior (see contract (i.e. individual beliefs, regarding terms group context below). of an exchange agreement between themselves and their organization). Breach of this contract has been shown to have an impact on several work-related outcomes, such as job satisfaction, organizational commitment, mistrust, and perfor- mance, among others (Zhao et al, 2007). The Behavioral Professional I 28 MONETARY AND tivation exerted higher effort. The Public Sector NON-MONETARY INCENTIVES Motivation questionnaire offers a measure of civ- ic-mindedness, comprising six dimensions: At- Both monetary and non-monetary incentives can traction to Policymaking, Commitment to Public affect the performance of policy professionals Interest, Social Justice, Civic Duty, Compassion (e.g. Ashraf et al 2014). Individuals can be posi- and Self-Sacrifice. tively and intrinsically motivated by public, peer, or supervisor recognition, and this can affect their persistency and/or productivity (e.g. Gauri et al, 2019; Grant 2008). Recognition can also be informal or intangible, for example, acknowledg- ing good work behavior. This is a particularly use- One’s Interpretation ful practice for improving motivation and perfor- mance on day to day and more mundane tasks This relates to how factors linked with the way (Montgomery et al, 2008). Relatedly, individual’s the individual processes information – such trust in his/her employer helps strengthen invest- as mental models and beliefs about how the ment in nurturing the relationship and can pro- world works, cognitive biases affecting the in- mote reciprocity (Coyle-Shapiro, 2002). terpretation of information, or availability of cognitive resources, among others – affect his/ her decision-making and behavior. (PERCEIVED) RELEVANCE OF WORK The literature highlights the importance of inter- MENTAL MODELS AND BELIEFS esting (and perceived importance of) work within Civil servant behavior can be influenced by deep- public bureaucracies as a motivating factor (e.g. ly internalized beliefs about how the world works, Wright et al, 2003). Moreover, field staff percep- also known as cultural schema or beliefs (DiMag- tions about the need for, and potential benefits gio, 1997). This includes the default associations, of a given program, confidence in their ability to categories, concepts, identities, prototypes, ste- deliver, and skill proficiency impacts fidelity of reotypes, causal narratives, and worldviews that implementation (Abry et al, 2013; Durlak & DuPre, we use to make sense of the world. These cultur- 2008). Fidelity of implementation is likely to af- al schema shape perceptions and filter the “facts” fect the program’s effectiveness, and the percep- that people believe, and affect decision-making tion of its success as a result. and service delivery (World Bank, 2015). For in- stance, teachers can exhibit unconscious biases towards students based on socio-economic cues, CIVIC-MINDEDNESS which affect their evaluation of students’ perfor- mance, learning ability and potential (World Bank, Studies have shown that civic-mindedness, or 2018). the individual’s desire to provide public service, is positively associated with individual and orga- nizational performance, job commitment, ethical behavior, among others (Ritz et al, 2016; Callen et COGNITIVE BIASES al, 2018). For instance, Banuri and Keefer (2015) found that individuals with greater pro-social mo- Policymakers must deal with the interpretation of The Behavioral Professional I 29 data, such as when making decisions on alloca- BANDWIDTH tion of funds based on data on program impacts. Taxing mental bandwidth, or availability of cogni- Yet, several individual cognitive biases have tive resources, can impact decision-making. Mak- been documented among policy professionals ing a series of difficult decisions, distraction and that come in the way of objective and impartial time pressure can constrain the ability to monitor decision-making and affect performance. For in- decisions and notice possible errors (Soll, Milk- stance, they are subject to confirmation bias driv- man, & Payne, 2016). Milkman (2012) showed en by ideological predisposition when interpret- that incidental uncertainty reduces persistence ing data, “variance neglect” of impact evaluation on difficult tasks, possibly by causing ego-deple- results, availability bias when assessing informa- tion and reducing self-control resources. It has tion, sunk cost fallacy with regard to funds allo- also been shown that high workload can exacer- cation, to mention only a few (Banuri et al., 2019; bate implicit biases and discrimination of teach- Battaglio et al., 2019; Vivalt & Coville, 2019;World ers towards students (Andersen & Guul, 2019). Bank, 2015). A recent study documented the presence of sev- eral of these in an experimental setup with man- agers and employees in the Italian public sector (Belle et al, 2018). Another study found that poli- cymakers overweight positive impact evaluation One’s Ability results compared to negative results (overconfi- dence) and do not consider variance of results This relates to how factors linked with the in- when updating (variance neglect) (Vivalt & Co- dividual’s skills – such as his/her knowledge, ville, 2019). In short, biases can affect how evi- non-cognitive or soft skills, and technical abili- dence is perceived in different ways, which can ty – affect their decision making and behavior. have large implications in a public policy context due to the potential weight of the decisions made on policy choices, design and implementation. KNOWLEDGE AND BELIEFS Skill level can attenuate effects of certain bias- Policy professionals are responsible for policy es that require avoidance of heuristic errors (e.g. decisions, implementation choices and opera- conjunction fallacy) (Baker et al, 2020). However, tionalization. Their tacit knowledge and beliefs many biases such as confirmation bias, sunk cost can influence attitudes towards beneficiaries and anchoring effects remain uncorrelated with and the information used for decision-making cognitive ability (Stanovich & West, 2008). There (Keulemans & Van de Walle, 2018; Rogger & So- is evidence that experience may play a role in mani, 2019). A recent study with public officials some cases. For instance, more experienced bu- in Ethiopia, showed that a large proportion of reaucrats tend to be more overconfident in as- policy officials make substantial mistakes about sessing their expertise (Liu et al, 2017). the basic conditions of local jurisdictions, and that these errors predict distribution of economic resources, career success and productivity (Rog- ger & Somani, 2019). On a slightly related note, Bergman et al (2019) showed that professionals displayed significant The Behavioral Professional I 30 misperceptions about take-up rates under auto- TECHNICAL SKILLS matic vs. opt-in enrollment. This is despite over- Skill level may improve accuracy in judgment whelming evidence on the positive effects of and decision-making (e.g. numeracy in the case defaults and illustrates how implementation spe- of interpreting numerical data) and skill profi- cifics that can heavily affect impacts of programs ciency also impacts fidelity of implementation but go unnoticed by policy professionals. When as mentioned above (Abry et al, 2013; Durlak & presented with evidence, they were neverthe- DuPre, 2008). However, technical skills are likely less willing to pay 144% more for the technology not to be sufficient for effective program imple- under automatic enrollment. mentation. Providers’ expectations, motivation, confidence in the ability to deliver and sense of self-efficacy also play a role on the extent to which individuals adhere to the prescribed im- NON-COGNITIVE SKILLS plementation practices (Durlak & DuPre 2008). Non-cognitive skills (or soft skills) include perse- verance (“grit”), conscientiousness, self-control, attentiveness, self-efficacy, resilience to adversi- ty, empathy, among others. In a comprehensive review of the literature, Kautz et al. (2014) provid- ed evidence that that non-cognitive characteris- tics are often as predictive of economic success as cognitive skills. The Big Five personality traits (i.e. Openness to Experience, Conscientiousness, Extraversion, Agreeableness, and Neuroticism) represent a commonly used taxonomy, and each trait comprises multiple facets and/or relates to different non-cognitive skills (for details see Kautz et al, 2014). Callen et al. (2018) recently showed that, in addition to incentives and public service motivation, personality traits/non-cogni- tive skills affect service public delivery (e.g. con- scientiousness was positively associated with doctors’ being at work during an unannounced visit). In another study, Kalaj, Rogger and Somani (2020) showed that locus of control, or the ex- tent to which individuals base success on their own actions, mediated civil servants’ time use at work and additional evidence suggest that locus of control may be associated with promotion op- portunities, rewards and motivation (Kay, Rogger & Sen, 2019). The Behavioral Professional I 31 GROUP CONTEXT Social Identity Normative Expectations Policy professionals’ social identities can influ- Normative expectations (sometimes referred ence their behavior towards beneficiaries from to as injunctive norms) refer to what we be- the in- vs. out-group. Biases and stereotypes lieve others think should be done/what is ap- related to out-group beneficiaries can lead to propriate to do (Bicchieri & Xiao, 2009).). In an discrimination in information provision and re- institutional context, there are often unwritten sponse quality by bureaucrats or field staff (e.g. rules on actions are permissible, mandatory, or White et al. 2015; Butler & Broockman 2011; Neg- prohibited, which guide public officials’ behavior, gers 2018). Group identity can also prompt moti- relationships with beneficiaries, and implemen- vated reasoning and affect decision-making (see tation process. Mangla (2015) notes that these biases subsection above) (Nørgaard, 2018). bureaucratic norms influence how officials enact their responsibilities, deliver services or engage with beneficiaries, and can impact agencies abili- ty to work effectively. In the context of corruption in a policy setting, research has documented the Empirical Expectations role that norms such as reciprocity, support of family, loyalty can play (Jackson & Köbis 2018; Social norms influence decision-making and Scharbatke-Church & Chigas, 2019). When cor- behavior, and whether the individual conforms ruption is pervasive, in-group pressure can make to norms depends on empirical and normative behaving honestly costly (World Bank, 2017). expectations regarding others in their refer- ence group. Empirical expectations (sometimes referred to as descriptive norms) refer to what we believe others to do (Bicchieri & Xiao, 2009). Individuals can have misperceptions about norms, and this can affect their behavior, includ- ing in the context of organizations. One such ex- ample is pluralist ignorance, whereby individuals underestimate the extent to which colleagues may share their reservations on a given subject and refrain from voicing their objective opinion (Westphal & Zajac, 2013). The Behavioral Professional I 32 Annex 2 Common behavioral pitfalls among policy professionals SELECTIVE PERCEPTION BIAS: BIAS BLIND SPOT: Tendency to fail to notice or more easily dismiss Tendency to perceive oneself as less biased than information that contradicts one’s views or beliefs. other people, which can lead to more easily ignor- Policy professionals’ behavior can be influenced ing the advice of peers or experts. by deeply internalized beliefs about how the world works, also known as cultural schema or beliefs. STATUS QUO BIAS: Preference for avoiding change and resistance to CONFIRMATION BIAS: adapt to new factors and circumstances. Tendency to search for, interpret, or favor informa- tion that support our pre-existing beliefs or ideo- IMPLICIT BIAS: logical predispositions. Mental associations that can lead to unintentional discrimination or stereotyping; AVAILABILITY BIAS: Tendency to think that examples that easily come SOCIAL IDENTITY: to mind – such as those based on what one has These can influence behavior towards beneficia- observed or heard about (vs. statistics and histori- ries from the in- vs. out-group (e.g. discrimination cal facts) – are more representative than in actual in information provision and response quality). reality. GROUP-THINKING: STEREOTYPE BIAS: Institutions are social structures, and individuals Beliefs about a given social group reflective of the will at times conform to what is perceived as the individual’s underlying prejudices or other inter- group view and refrain from providing opposing nal motivations. Typically centered around most ideas to avoid disagreement; distinctive features of the group, these can result in distorted emotions and perceptions of group EMPIRICAL EXPECTATIONS: members (e.g. Bordalo et al., 2016). Sometimes referred to as descriptive norms, these refer to what we believe others to do (Bicchieri & Xiao, 2009). Misperceptions about norms can negatively affect behavior (e.g. pluralist ignorance, The Behavioral Professional I 33 whereby individuals underestimate the extent to which colleagues may share their reservations and refrain from voicing their objective opinion). NORMATIVE EXPECTATIONS: Sometimes referred to as injunctive norms, these refer to what we believe others think should be done/what is appropriate to do (Bicchieri & Xiao, 2009). In an institutional context, there are often unwritten rules on actions are permissible, man- datory, or prohibited, which guide public officials’ behavior, relationships with beneficiaries, and implementation process. The Behavioral Professional I 34 Stay Connected eMBeD@worldbank.org #embed_wb worldbank.org/embed bit.ly/eMBeDNews