University of Copenhagen

Transcription

1 university of copenhagen University of Copenhagen Behavioral Public Administration: Combining Insights from Public Administration and Psychology Grimmelikhuijsen, Stephan; Jilke, Sebastian; Olsen, Asmus Leth; Tummers, Lars Publication date: 2015 Citation for published version (APA): Grimmelikhuijsen, S., Jilke, S., Olsen, A. L., & Tummers, L. (2015). Behavioral Public Administration: Combining Insights from Public Administration and Psychology. Download date: 01. Jul. 2016

2 1 Behavioral Public Administration: Combining Insights from Public Administration and Psychology Stephan Grimmelikhuijsen, Utrecht University; Sebastian Jilke, Erasmus University Rotterdam; Asmus Leth Olsen, University of Copenhagen; Lars Tummers, Erasmus University Rotterdam; Authors are listed in alphabetical order, each author contributed equally. Words: 9,059

3 2 Abstract We propose behavioral public administration as a designated subfield in public administration which explicitly deals with the integration of theories and methods from psychology into the study of public administration. We discuss how scholars in public administration currently draw on both methodological and theoretical innovations in psychology and point to research questions in public administration which could benefit from further integration. Behavioral public administration cannot, and should not, replace conventional public administration, but it is complementary to it. Importantly, behavioral public administration represents a two-way street in which public administration scholars use theories and methods from psychology, and psychologists, in turn, learn from our setting of political-administrative contexts to refine their theories and methods. Finally, we propose four principals to advance the agenda of behavioral public administration to narrow the gap between public administration and psychology with the aim of advancing our current understanding of public administration theory and practice. Keywords: Behavioral public administration, Psychology, Behavioral Sciences, Experiments

4 3 Introduction Eminent public administration scholars, such as Herbert Simon and Dwight Waldo, have repeatedly stressed the importance of psychological research for the study of public administration (Simon 1947a, 1965, 1979; Waldo 1948, 1965). Simon (1947b, p. 202) even stated that [f]or the man who wishes to explore the pure science of administration, it will dictate at least a thorough grounding in social psychology. Since these early calls to integrate both fields of study, and despite the inherent interdisciplinary nature of public administration (Mosher 1956; Raadschelders 2011), public administration scholars have partly neglected theories and methods from psychology (Jones 2003). If we look at neighboring disciplines such as economics, political science, or management studies, psychological research had a much more profound impact within these fields of inquiry. This is reflected in the emergence of psychology-informed subfields such as political psychology (see for instance McDermott 2004), behavioral economics (e.g., Mullainathan and Thaler 2000) and industrial and organizational psychology (e.g., Cascio 1995). Yet, recently there has been an increase of public administration studies borrowing and extending theories from the field of psychology. This includes studies on core public administration topics such as public leadership and motivation (Bellé 2015), transparency (Grimmelikhuijsen and Meijer 2014), public service competition and choice (Jilke 2015), performance information (James 2011; James and Moseley 2014; Olsen 2015a) and trust of civil servants (Van Ryzin 2011). At the same time, the methodological toolkit of public administration scholars is becoming more informed by developments from psychology, for instance, by using experimental methods (Margetts 2011) and measurement techniques (e.g. Kim et al. 2013; Tummers 2012). But this is not a one-way street. Some psychologists are explicitly connecting their theories to the field of public administration, among else via articles in Public Administration Review (Bakker forthcoming, Tybur and Griskevicius 2013; Wright and Grant 2010). These developments may mark the starting point of the emergence

5 4 of a psychology-informed subfield of public administration, in which public administration scholars use theories and methods from psychology and psychologists, in turn, learn from our setting of political-administrative contexts to refine their theories and methods (Olsen 2015b). In this article, we aim to deepen the dialogue between public administration and psychology by outlining a distinct subfield in public administration which integrates both fields of study: behavioral public administration. We describe behavioral public administration as a subfield that is characterized by the interdisciplinary analysis of public administration from the micro-perspective of individual behavior and attitudes by drawing upon recent advances in our understanding of the underlying psychology and behavior of individuals and groups. This definition has three main components: 1) it rests on a microlevel focus (i.e. (groups of) citizens, employees and managers within the public sector are the unit of analysis); 2) it studies how these people behave and form attitudes; and, most importantly, 3) it does so by integrating insights from psychology and the behavioral sciences into the study of public administration. In short, a behavioral public administration studies the behavioral foundations of public administration through theories developed in psychology and the behavioral sciences more broadly (Jilke 2015). This article outlines a behavioral public administration by presenting and connecting four analyses. First, we review the historical background on a closer bond between public administration and psychology. Second, we draw lessons from the neighboring fields political psychology and behavioral economics, which have witnessed a similar development in the recent past. Thirdly, we show using recent examples of the application of psychological theories and methods in public administration how this subfield can contribute to nuance, support or criticize grand theories in our field. Fourth and finally, we conclude with an agenda comprising four guiding principles to further develop a behavioral public administration.

6 5 A Reemerging Call for Integrating Public Administration and Psychology Before outlining how public administration research may benefit from psychology, it is worth showing how this idea is rooted in the ideas of early public administration scholars. The point of departure for our argument is therefore to take a closer look at historical calls for integrating insights from psychology into public administration. This exercise reveals that we do not have to search long or in obscure corners of our field: Discussions about the crossfertilization of the two disciplines were central to many seminal figures in public administration scholarship. The most obvious early attempt is found in Herbert Simon s seminal work Administrative Behavior (Simon 1947a). Today, Simon's scholarship is mostly credited with introducing concepts of bounded rationality and satisficing into the study of public administration (e.g. Meier 2015). While these concepts clearly draw on insight from cognitive and social psychology, they are just examples of how Simon envisioned a much more general and tight integration between the two fields. In his Nobel Prize speech of 1978 he cited how his 1947 book grew from the conviction [ ] that decision making is the heart of administration, and that the vocabulary of administrative theory must be derived from the logic and psychology of human choice. (Simon 1978, p. 353). Simon reiterated this view in an exchange with Robert Dahl in Public Administration Review over the fundamentals of public administration as a science: For the man who wishes to explore the pure science of administration, it will dictate at least a thorough grounding in social psychology. (Simon 1947b, p. 202). This sentiment reflects the idea that public administration is subordinate to psychology, as administrative decision-making must be studied as a special case of the many forms of decision-making studied by psychologists: The research worker in administration must consider himself not merely a person whose work is related to social psychology, but a person who is a social psychologist concentrating in a particular special area of human behavior. (Simon 1947b, p. 203). In other words, one

7 6 of the founding fathers of contemporary public administration saw the field as ideally an applied subfield within social psychology. However, Simon also noted that public administration cannot merely be a passive user of psychology, but must aim to be [ ] as capable of contributing as it is of accepting (Simon 1947b, p. 203). A decade later Simon still saw a great distance between public administration and psychology (Simon 1955). Again, he recognized that psychology also could learn something from public administration and thus a [ ] a marking stone placed halfway between might help travelers from both directions to keep to their courses (Simon 1955, p. 100; see also Olsen 2015b). In other words, Simon envisioned a two-way street between the two fields. But Simon was not the only eminent public administration scholar interested in psychology. In The Administrative State published in 1948, Dwight Waldo also discussed the connection between public administration and psychology. He noted how psychologists see [ ] that man is in small part rational (Waldo 1948, p. 25), but rather motivated by emotional drives and urges. However, he also notes how [p]ublic administration has been little touched by such ideas (Waldo 1948, p. 25). Later on Waldo (1965) evaluated the extent to which psychological insights had penetrated public administration. While he noted that psychology had entered public administration through various minor channels, the overall impression was still at the time that, [ ] a rank amateur in Psychology can quickly identify theories and techniques no yet brought into our professional range (Waldo 1965, p. 5). Calls for integrating insights from psychology into public administration extend beyond Simon and Waldo. David Truman (1945, p. 69) makes the case for adopting public opinion research in public administration and points out that [a]dvances in psychology may one day permit planned variation in organizations according to the measured qualities of personalities in or available to them. In a similar optimistic vein, Frederick Mosher (1956, p. 178) discusses the role of public administration in relation to other social sciences and points to the [ ] inadequate collaboration, cooperation, and interaction between them -

8 7 including in particular public administration and the field of psychology. Along the same lines scholars in the 1950s and 1960s argued for a much tighter integration between the two fields (e.g. Honey 1957; Rosenthal 1959; Verba, 1961). The abovementioned calls illustrate how ideas about the connection between public administration and psychology were part of early discussions about the foundations of contemporary public administration. These discussions were not about defining the boundaries between the two fields, but rather focused on pointing to the obvious intersections. Simon and Waldo s repeated disappointments concerning the far distance between psychology and public administration seem to ring true to some extent even today. As Jones (2003, p. 395) has argued: Most people who study politics and government care little about the fine details of the specifics of human cognition. However, as already shown in the introduction, there seems to be more psychologically-informed public administration research that meets the eye, although a psychology-informed subfields has not yet developed. This will be elaborated upon in Section 4. But first, we will shift focus to two neighboring disciplines (political science and economics) that have successfully integrated psychology into their discipline and developed psychologically-informed subfields. The Role of Psychology in Neighboring Disciplines If we look at neighboring disciplines such as political science and economics, psychological research has had a profound and lasting impact within these fields of study. This is reflected in the emergence of the psychology-informed subfields political psychology and behavioral economics. The examination of these subfields provides possible guidance for developing the subfield of behavioral public administration.

9 8 Political Psychology The study of individual-level political behavior is one of the key topic areas of contemporary political science. But while the behavioral-turn in political science of the 1950s and 1960s (see Dahl 1961) has shifted the focus of political researchers to the empirical analysis of political phenomena at the individual-level, the comprehensive integration of psychological theories into the study of political behavior appeared to have happened decades after. This can be exemplified by the emergence of the interdisciplinary subfield of political psychology, which gained most of its prominence in the period since the 1980s (Lavine 2010). Yet seminal works can be traced back to the 1930s, such as Harold Lasswell s Psychopathology and Politics (1930) (see Ascher and Hirschfelder-Ascher 2005). Contemporary political psychology emphasizes the intertwined nature between politics and psychology, whereby the study of political psychology can be regarded as an interdisciplinary field that studies political behavior from a psychological perspective (e.g. Sniderman and Tetlock 1993; Lau and Redlawsk 2006). The field of political psychology has grown tremendously in past years, reflected in the establishment of its own journals, such as Political Psychology, a great share of political psychological studies appearing in the flagship journals of political science (for instance Kuklinski et al. 1991; Petersen and Aarøe 2013). The two editions of the Oxford Handbook of Political Psychology (Sears, Huddy and Jervis 2003; Huddy, Sears and Levy 2013) neatly illustrate this trend. Moreover, psychological research s success in political science - which is often contrasted with rational choice approaches - is reflected in various fields within the discipline, such as the study of voting behavior (Gerber, Green and Shachar 2003), political attitudes (Taber and Lodge 2013), but also within the area of international relations (McDermott 2004). An example of the application of psychological research into political behavior is the study of the theory of motivated reasoning (see for instance Lodge and Taber 2006). Put

10 9 simply, motivated reasoning is a type of cognitive dissonance reduction that puts forward the idea that people try to conform to their prior held values and beliefs when making decisions, or attitudes (see Festinger 1957). According to this theory, people will hold upon false beliefs and interpret contrary information in a way that is most favorable to their prior values and beliefs. In this notion, Sloothuus and de Vreese (2010) study how political sponsorship of issue frames (i.e., party statements about certain policy areas within newspapers) using two population-based survey experiments. They find that citizens tend to vary their support of the very same issue frame in accordance to its political sponsorship all other things being equal. In other words, people respond differently to information about policy issues when it is presented by the party they voted for, than if the frame was sponsored by another party. These findings are explained by arguing that people try to ensure conformity with their prior vote choices by adjusting their support to policies. Political psychology has also employed insights from social psychology to explain topics such as leadership behavior and group processes in decision-making (e.g. Golembiewski and Miller 1980; t Hart, De Jong and Geveke 1993). An example of how group behavior affects individual political thought is a study by Erisen and Erisen (2012) who find that close-knit social networks could create social bubbles that limit how one communicates with others and reasons about politics, eventually declining the quality and broadness of policy-relevant thinking. Another example regards the concept of groupthink, which builds on psychological theories about how individuals make decisions in groups (Janis 1972): a strong desire for harmony and consensus, group insulation, and lack of impartial leadership can cause a group of talented and experienced political decision-makers to make disastrous collective political judgments ( t Hart 1994). In sum, this section made clear how in the neighboring field of political science the integration of social and cognitive psychological insights has led to valuable theoretical advances in the areas of, for instance, decision-making and motivated reasoning. The next

11 10 section explores behavioral economics, which has also integrated psychological insights, but has developed in a quite different direction. Behavioral Economics Psychological research nowadays also has a prominent place within behavioral economics. A good example of this integration is psychologist Daniel Kahneman who received the 2002 Nobel Prize in economics. Being a psychologist by training and highly regarded in both fields, he, and his long-time collaborator Amos Tversky, became founding fathers of contemporary behavioral economics by integrating insights from psychological decisionmaking theory to individual's market decisions and judgments. Kahneman (2003, p.1449) noted that Tversky and he primarily viewed their work to contribute to psychology, but that they were [ ] drawn into the interdisciplinary conversation by economists who hoped that psychology could be a useful source of assumptions for economic theorizing. Their body of work displays the limitations of neoclassical economics (e.g. Kahneman and Tversky 1979; Tversky and Kahneman 1974; 1981). Indeed, these and other insights from psychological research about the limited selfishness, cognitive biases and bounded rationality of individuals has led economists to reconsider the cornerstone of their conceptual frameworks, the Homo Oeconomicus (for overviews see Rabin 1998; Sent 2004). In this realm, behavioral economics emerged as an independent subfield. It can be defined as "[ ] the combination of psychology and economics that investigates what happens in markets in which some of the agents display human limitations and complications." (Mullainathan and Thaler 2000, p. ii). In other words, behavioral economists study individual's market behaviors by placing a major emphasis on how agents deviate from the neoclassical conception of the rational man. Mullainathan and Thaler (2000) put forward three categories for agents' deviations: 1) bounded rationality and limited cognitive abilities, 2) bounded willpower, and 3) limited self-interest and reciprocity.

12 11 Applications of behavioral economics span across traditional economic subfields such as macroeconomics, labor economics, or finance. Yet, in the area of microeconomics, behavioral economics seems to be most dominant (Camerer, Loewenstein and Rabin 2004). An interesting example can be drawn from the area of charitable giving. Ariely and colleagues' (2009) studied people's responses to monetary incentives using psychological motivation theory. Neoclassical economic theory would assume that monetary incentives steer agents to behave more prosocially. Yet, building upon self-determination theory (Deci and Ryan 1985), motivation crowding (Titmuss 1970; Frey 1997) puts forward the idea that increasing extrinsic types of motivation (any type of reward that is associated with doing good), can lead to undermine individual's intrinsic (the value of doing good per se), and - eventually - image motivations (the degree to which individuals are motivated by others' social approval). Based on these assumptions, Ariely and colleagues (2009) assess the potential adverse effects of extrinsic rewards on the relationship between people's image motivation and their prosocial behavior, using a set of experiments. They find that private monetary rewards indeed crowd-out image motivation; When receiving financial rewards, the effect of doing good because of social approval diminishes. This study highlights the fruitful application of psychological theories in the realm of economic behavior. Using the experience of political psychology and behavioral economics What can we learn from how these neighboring disciplines deal with insights from psychology? There are differences in how they integrated psychological research into their mother-disciplines. For instance, behavioral economics predominantly shows that individual behavior in markets deviates from the assumption of the Homo Oeconomicus, while political psychology uses psychology to explain individual s political behavior instead of solely examining institutions at the macro-level. But both disciplines are similar in that they show

13 12 that there is much to learn from psychological theories and have integrated this into their own field of study. As stated, public administration has not developed a similar psychologically informed subfield as found in these fields. In fact, our own Herbert Simon is credited in both political psychology (Sniderman and Tetlock 1993) and behavioral economics (Simon 1978; Kahneman 2003) as a central source of inspiration. And the list of historical calls for a tighter integration confirms that many scholars have seen and acknowledge the links between the two disciplines. The task ahead and the scope of this paper is therefore not to understand why opportunities were forgone, but rather how current and future research in public administration and psychology can pave the way for an interdisciplinary subfield of behavioral public administration. Behavioral Public Administration: Theory and Methods In this section we shift our focus from how other fields have integrated psychological theories into new subfields, towards our own field. We will show how behavioral public administration can be beneficial for theory and practice of public administration, and how it can be further developed along theoretical and methodological lines. Theories for Behavioral Public Administration A behavioral public administration can be beneficial for public administration to, firstly, connect grand macro-level theories with their micro-level foundations. Thereby, a behavioral public administration can verify, falsify or nuance claims made at the macro-level and thus refine institutional theories through the systematic testing of their microfoundations (Jilke 2015: Stoker and Moseley 2010). We will exemplify this by using examples to show how studies that we would classify belonging in the realm of a behavioral public

14 13 administration can verify and nuance macro-level theories on the impact of 1) transparency on trust, and 2) representative bureaucracy on trust, fairness and performance. Transparency The first example regards a micro-level test of institutional theories of government transparency. Government transparency is said to strengthen citizen trust and legitimacy (Hood and Heald 2006). Being transparent shows that governments have nothing to hide and provides citizens with information that helps them to better understand public processes of decision- and policy-making. However, the positive impact of transparency on trust might not be as straightforward as claimed and recent research suggests that the effects of transparency on citizen attitudes depends on contextual factors (e.g. Worthy 2010; Grimmelikhuijsen et al. 2013). For instance, an experimental study by De Fine Licht (2014) tests the micro-level underpinnings of the relationship between government transparency and trust. Doing so, De Fine Licht uses the psychological concept of taboo trade-offs (e.g. Fiske, Page and Tetlock 1997) to show how citizens respond more skeptically to government decision-making transparency in areas that relate to human life and death, compared with less controversial areas. According to this theory, human life and well-being are considered sacred and cannot be traded off against secular values, such as money. Government transparency can expose decision-making of these trade-offs. Participants who were exposed to decisionmaking about a taboo trade-off, such as traffic security, perceived the decision-maker as less legitimate than non-taboo decisions (such as about parks and recreation). It illustrates that government decisions about such trade-offs will therefore encounter much more resistance than trade-offs that do not violate this taboo (De Fine Licht 2014). This example from transparency research exemplifies how the integration of psychological theories can explain individual behavior in a public administration context, and more importantly how it can

15 14 nuance macro-level theories. It shows that transparency does not automatically lead to higher levels of trust as assumed by institutional theory, but that it is conditional on whether the decisions concern controversial areas. Representative bureaucracy A second example tests some of the microfoundations of the theory of representative bureaucracy. The idea behind this theory is that a representative composition of the workforce of a bureaucracy in terms of gender, racial and ethnic diversity can promote democracy in various ways. For instance, Theobald and Haider-Markel (2009) have highlighted that representation makes bureaucracies more responsive and increases government accountability. One of the key micro-level assumptions underlying representative bureaucracy is that of symbolic representation. According to them (2009, p. 410) [...] symbolic representation works cognitively on the audience of those who belong to a group that is to be represented. With symbolic representation, then, attitudes and outcomes can change without any purposeful actions taken by the representatives other than holding a government office or position. This means that having a more representative workforce should have beneficial effects on how an audience perceives it. To test this microfoundation of representative bureaucracy theory, Riccucci, Van Ryzin and Lavena (2014) designed an online survey experiment. In this experiment, hypothetical scenarios of police units were randomly assigned to four experimental groups. The scenarios varied in their levels of displayed performance and representativeness (i.e., gender composition). For instance, in one scenario the organizational unit consisted of a representative workforce of 4 males and 6 females, whereas in another scenario this was very unequally distributed (9 males and 1 female). After the experiment, participants were asked about the trustworthiness, fairness and performance of the police unit. The findings of this study show that a representative workforce is evaluated more positively on trust, fairness and

16 15 performance. This effect was even stronger for female participants. It confirms one of the key micro-level assumptions of representative bureaucracy theory, namely that symbolic representation causes the audience to feel that they are being represented, in this case women. In sum, these two examples show how a behavioral public administration can contribute to nuance (transparency case) and verify (representative bureaucracy case) the microfoundations of macro-level theories. It also shows that a behavioral public administration cannot, and should not, replace conventional public administration research, but that it is complementary to it. Methods for Behavioral Public Administration Besides its theories, psychological research can also contribute to the methodological development within public administration. We emphasize two methodological contributions that research in psychology, and the behavioral sciences more broadly, can make to how we conduct research within behavioral public administration. The first contribution focuses on the use of experimentation as a distinct logic of inquiry, while the second places its emphasis on measurement. Experimentation Public administration scholars have only recently started using experiments more frequently; Anderson and Edward (2014) state that in recent years more experimental research has been published in public administration journals than during the last decades. For instance in 2013, already nine experimental studies were published, which is more than the entire 1960s, 1980s and 1990s and it steadily increases. That being said, public administration today is far from an experimental science at the same scale as psychology where thousands of experiments are published each year. However, we must note that arguments for methodological choices depend partly on the research problem at hand and partly on the philosophical or

17 16 paradigmatic preferences of the researcher (Haverland and Yanow 2012). Hence, we argue that experimentation is useful for, and inherently experimental in, every design science - such as public administration (Shangraw, Crow and Overman 1989) - but we are not favoring an outright replacement of other social science techniques (such as participant observation or interviews) by an experimental logic. What makes experimentation a useful method? When comparing this technique of social inquiry to more conventional ways of gathering social data, we see that nonexperimental methods are potentially prone to a range of methodological shortcomings that stem from the endogenous nature of the relationships they investigate. This endogeneity is a result of a range of different biases, such as omitted variable bias (the presence of confounding factors), selection bias (results stem largely from the selection of analytical units) or simultaneity bias (i.e. reverse causality) (e.g. Angrist and Pischke 2009). Experimental designs provide a clear cut solution to the problem of endogeneity and can present a benchmark for the internal validity of social research. Through the process of randomization to treatment and control conditions, scholars are able to examine the counterfactual effect that is what would have happened without that treatment of a specific intervention. The benefits of an experimental logic can be shown using a core public administration topic: red tape. Red tape can be defined as rules, regulations and procedures that entail a compliance burden without advancing the legitimate purposes they were intended to serve (Bozeman 2000, p.12). In other words, these rules are burdensome, while providing no added value. Here, an experimental approach could be helpful. Comparisons between control and treatment groups can determine what would have happened in the absence of red tape (or lower degrees thereof), while the process of randomization ensures the statistical equivalence of experimental subjects. An interesting example here is the work of Scott and Pandey (2000). They used an experimental simulation to address the impact of red tape on

18 17 bureaucratic behavior, such as recommending financial assistance for a client (for instance food and clothing allowance). The findings showed that increasing levels of red tape produce a reduction in benefits recommended for clients. A possible implication is that comparable clients may be treated differentially only as a function of the level of red tape involved in the benefit determination process. This illustrates how public administration scholars can develop important practical and scientific insights using experiments. However, we argue that the value of experimentation for the study of a behavioral public administration lies in its distinct logic of inquiry, rather than in the mere conduct of an experiment. Thus the same logic of inquiry may also be applied to non-experimental data by developing a causal identification strategy via exogenous variation variation that comes from outside the theoretical and empirical model. This is what has been labeled a designbased approach towards experimentation (Dunning 2012; see also Angrist and Pischke 2009). An example of exogenous variation that could be exploited for the causal identification of outcomes are, for example, the historical courses of state borders such as the ones from the Habsburg Empire which spread across, among others, Austria, Belgium, Croatia, Czech Republic, and many others because living nowadays within a short distance to former country borders (on both sides) can be regarded as-if randomly allocated. Thus using the former Habsburg borders as exogenous variation, Becker et al. (2014) test whether the longgone Habsburg Empire with its well-regarded public administration had a casual effect on today s levels of trust and practice of corruption, via a so-called regression discontinuity design. They find that within the areas that used to belong to the former Habsburg Empire, higher levels of trust and a lower prevalence of corruption in courts and police have been developed. Drawing upon such exogenous variation enables researchers to overcome problems of endogeneity by establishing a causal linkage between two variables of interest. The experimental logic thus displays a benchmark which can be used to answer causal research questions.

19 18 However, it needs to be noted that experimentation has also its own potential pitfalls and downsides. For instance, experimental designs can lack external validity. Yet, direct and conceptual replications, rigorous systematic reviews and meta-analyzes of experimental data can help to increase the external validity of experimental results. Here, a behavioral public administration can learn a great deal from research in psychology and psychology-informed subfields, such as behavioral economics and political psychology. A related concern is that experiments are often very abstract. One can question whether this abstractness works for public administration research. Bozeman and Scott (1992, 309) for instance, argue that experiments in public administration research should require attention to mundane reality. This means there should be sufficient levels of contextual realism in experiments in order to be able to use the results in public administration practice. As a result, public administration experiments are likely to be more complex than the elegant but often abstracted and stylized social psychological experiments. Another concern is the costs and time needed to properly conduct experiments, especially when considering field experiments and potentially small effects, requiring large samples (Gerber and Green 2012). Measurement Paying attention to measurement bias is an additional methodological opportunity that arises when working towards a greater integration of psychology within the study of public administration. It is promising to see that lately issues of measurement has been taking up by various public administration scholars. For instance, articles have been published regarding priming or order effects within scales (Van de Walle and Van Ryzin 2011), cross-national measurement equivalence (Jilke, Meuleman and Van de Walle 2014, Kim et al 2014), social desirability bias (Kim and Kim 2013), common source bias (Favero and Bullock 2015) and the use of language in item wording (Feeney 2012). However, it seems that the measurement quality of public administration research can be strengthened further. For instance, a recent

20 19 review of the major public administration journals using the Total Error Framework by Lee and colleagues (2012) showed that many public administration studies are prone to various measurement errors, such as questionnaire wording, social desirability bias, overly complicated questions, data coding and estimation. They also found that the studies often fail in reporting possible sources of error. The value of sound measurement can be illustrated using an example on scale development. In general, public administration researchers have long carried out their quantitative studies without the use of validated scales. Many scholars solve this by developing their own, ad hoc, scales, or use single item measures. For instance, in a recent 30 year literature review on coping during public service delivery Tummers et al. (2015) found that no studies used validated scales. Instead of just writing up one or more items to measure the construct, it can be beneficial to develop scales using scale development techniques. In his work on scale development, DeVellis (2012, p.10) notes that Poor measurement imposes an absolute limit on the validity of the conclusions one can reach. Although in general public administration scholars do not employ scale development, there are exceptions, such as work on policy alienation (Tummers 2012), administrator s trust in citizens (Yang 2005), or public service motivation (Perry 1996; Kim et al. 2014). Hence, next to incorporating an experimental logic, a greater emphasis on measurement can help a behavioral public administration in making valid inferences which are also comparable across studies and contexts. This way a body of knowledge that is subject to continuous empirical investigation and adjustment can be developed to build rigorous scientific evidence of both, theoretical and practical pedigree. However, we acknowledge that using measurement techniques such as these have their own potential pitfalls. For instance, especially when analyzing new topics, validated scales may not exist. In addition, some items from validated scales may be too generic for a specific research topic, resulting in meaningless questions for respondents. This could draw

21 20 researchers away from such topics, and keep analyzing topics which have been studied extensively (such as public service motivation). This may hamper scientific progress. Furthermore, including social desirability checks would be difficult, for instance because questionnaires are already long. We would argue that in such cases (and similar ones), researchers should make an explicit trade-off between empirical rigor and scientific innovation. For instance, they can choose to lean towards innovation in the start of the research process (i.e., when studying a new topic area) and towards developing more scientific rigor when first studies seem promising. An Agenda for Developing a Behavioral Public Administration In this article, we have argued for a tighter integration of psychology and public administration under the banner of behavioral public administration. We defined it as a subfield that is characterized by the interdisciplinary analysis of public administration from the micro-perspective of individual behavior and attitudes by drawing upon recent advances in our understanding of the underlying psychology and behavior of individuals. This article showed how our call for developing such a subfield is (1) rooted in early public administration scholarship, (2) how we can learn from how political psychology and behavioral economics have successfully integrated psychological insights, and (3) how current developments in public administration research elicit the added value of behavioral public administration. It is important to stress that behavioral public administration is intended as an interdisciplinary field, being a subfield of public administration. As with any other subfield its role is to function as a transportation belt between some specialized topic of research and a broader discipline. We have argued that the role of behavioral public administration could be to make sure that public administration research has an ongoing dialogue with psychology on

22 21 the theories and methods relevant to administrative-political settings. Our plea is therefore one of pluralism and greater cross-disciplinary cooperation. As we have outlined, there are already numerous examples of recent public administration research which fit under the umbrella of behavioral public administration. These examples are found in certain areas of public administration (for instance red tape, motivation, transparency and representative bureaucracy), but much more sporadically in others (such as public management reform, agenda setting, or network governance). With behavioral public administration there is an opportunity to make a connection between the psychological insights already drawn upon across different research topics and at the same time outlining a framework for extending the application of psychology to other areas of public administration. What could scholars do to further develop behavioral public administration as a mature subfield of public administration and psychology? We envision the following four principles to further develop and manifest a behavioral public administration: 1) extending behavioral public administration to more public administration topics, 2) methodological advancement, 3) strengthening behavioral public administration as a mature subfield, and 4) increasing value for public administration practice. 1. Extending behavioral public administration to more public administration topics Some areas of public administration research have already witnessed an increase in research that could be placed under the banner of behavioral public administration. For instance, public leadership, policy decision-making, government transparency, representative bureaucracy, public service motivation, performance information and red tape. However, other substantive subfields could also benefit from a stronger connection with psychology. Without aiming to be exhaustive, important areas such as e-government, network governance, street-level bureaucracy and accountability are fields that are suitable for a behavioral public administration approach.

23 22 One of the major foci of e-government research, for instance, is the effects of e- service delivery on citizen behavior and attitudes (West 2004). This can be suited for behavioral public administration as it concerns a micro-level focus on individual citizens and how they form attitudes and behavior in a public sector context. In network governance and network effectiveness research identifies three levels of analysis: community, network, and organization/participant levels (Provan and Milward 2001). Especially at the latter level, psychological theories about collaboration and competition could help to better understand how and people in networks (do not) work together. One might think of theories on group think in decision-making (Janis 1972, see for an application of this in public administration t Hart 1994) and group conflict (Curseu and Schruijer 2010). The study of accountability could as well benefit from insights from experiments done by psychologists. Most notable, Philip Tetlock (1983) carried out experiments and showed how individuals engage in more complex information processing and elaborate justifying behavior if there is a hostile audience. The application of these theories could be very useful for public organization who often operate in a hostile media environment. 2. Methodological advancement Next to using behavioral approaches in more areas of public administration, behavioral public administration can also help in the advancement of other research methods such as field experiments, laboratory experiments and methods focused on causal inference more widely. It could integrate methods and research standards from psychology and psychology-informed subfields such as political psychology and behavioral economics. Indeed, the study of microlevel behavior of people within the public sector provides a fertile testing ground for the utilization of (quasi-)experimental techniques, and their measurement. Such a methodological broadening could contribute to increasing the credibility of usable knowledge generated within the discipline.

24 23 But how can we foster this methodological broadening? One possibility could be through an increased experimental training within graduate schools. Here, collaborations with psychology departments, or interdisciplinary method schools are interesting possibilities. Moreover, interdisciplinary collaborations between scholars may result in further learning about the use of sophisticated research techniques, for example how to conduct, analyze and report on experiments. Thereby, the utilization of experimental techniques and measurement methods which are considered state-of-the-art in psychology, could contribute to a behavioral public administration which produces rigorous and relevant findings. 3. Strengthening behavioral public administration as a mature subfield Every field and subfield has a point of departure in terms of more or less well-established ideas about suitable theories and methods for a specific research question. This process also entails that some theories or methods are ruled out or ignored altogether at the very onset of a research project. Taken together, the integration of psychological theories and methods into public administration rests on a fundamental inter-disciplinary principal for the development of public administration research. What does this look like in practical research? It first of all implies an open attitude toward robust theories and methods that are not initially developed for studying research questions that arise in an administrative setting. Specifically, this entails that a research project in behavioral public administration not only draws on the frontier of the best available knowledge within public administration but also is informed about state-ofthe art knowledge from psychology. That is, behavioral public administration does not study a specific questions in a vacuum but embraces at face-value the validity and relevance of the scientific frontier in psychology until tested further in a political-administrative setting. 4. Increasing value for public administration practice

25 24 Last but certainly not least, behavioral public administration can be beneficial for practitioners, such as policy makers, public managers and public professionals. The gap between research and practice in public administration has been intensively debated. Some scholars are quite critical about the applicability of public administration research to practice. For instance, Bogason and Brans (2008, 92) comment that The weak reception and application of public administration theory in practice suggests that the community of public administration academics may still be producing knowledge whose legitimacy and usefulness is questionable. On the other hand, other scholars are more nuanced concerning the theorypractice relationship. For instance, O Toole (2004, p.312) notes that: Expecting some theory, any theory, to translate simply into a clear and uniform body of knowledge suitable for all such customers is to expect far too much. The theory practice nexus is not a simple link in some translation belt from thought to action. Although he is not as pessimistic as some other scholars on the theory-practice relationship, he does acknowledge that the theorypractice relationship needs to be improved. Here behavioral public administration can help by developing usable knowledge. Perry (2012, p.479) argues that usable knowledge should a) meet the priorities of users and b) is research of high quality. Hence, behavioral public administration should develop usable knowledge by conducting high quality research on topics which are valuable for practitioners. Furthermore, scholars can at times explicitly try to connect their research to ongoing debates in practice, such as regarding ethical decision-making of politicians and managers. Furthermore, they can try to set policy agenda, as has been done by behavioral economist Richard Thaler and Cass Sunstein in their study on nudges (Thaler and Sunstein 2008). Next to writing scholarly articles and books, or even books and op-eds for the general public, scholars can also more directly try to help governments. Richard Thaler himself for instance became actively involved in the Behavioral Insights Team (or nudge unit), an organization set up to apply behavioral economics and psychology to improve government policy in the

26 25 United Kingdom. Another possibility would be that PhD-students work on a Ph.D. project which is both scientifically and practically valuable, and in order to improve the connection work for a number of days in the governmental organization, and the other days at the university. In these and other ways, scholars interested in behavioral public administration can try to develop usable knowledge and help to apply it. Concluding, to paraphrase Herbert Simon (1955, p. 100), we hope that behavioral public administration may develop into a marking stone placed halfway between psychology and public administration, thereby helping scholars from both directions to learn from each other. References t Hart, P Groupthink in Government: A Study of Small Groups and Policy Failure. John Hopkins University Press. t Hart, P., de Jong, P. and H.G. Geveke Sociaal-psychologische perspectieven voor onderzoek van beleid en bestuur. Beleidswetenschap. 7(1):3-33. Anderson, D. M., and Edwards, B. C Unfulfilled Promise: Laboratory experiments in public management research. Public Management Review, online first. Angrist, J.D. and J.-S. Pischke Mostly Harmless Econometrics: An Empiricist s Companion. Princeton University Press. Ariely, D., A. Bracha and S. Meier Doing Good or Doing Well? Image Motivation and Monetary Incentives in Behaving Prosocially. American Economic Review, 99(1): Ascher, W. and B. Hirschfelder-Ascher Revitalizing Political Psychology: The Legacy of Harold D. Lasswell. New York: Psychology Press. Bakker, A.B A Job Demands-Resources Approach to Public Service Motivation.