Toward a theory of behavioral public administration

Introduction

The rise of behavioral economics for administrative purposes is upon us. Behavioral policy departments have taken hold worldwide, starting from the British and Australian governments and now reaching into East Asia. There is growing recognition that the grounds of public policy should be open to alternatives to the standard economic rationalities in light of empirical findings of differing behaviors (Benartzi et al., 2017; John, 2013; Oliver, 2015). The natural affinity between the disciplines goes back to Herbert Simon, a pioneer of the less dismal science of behavioral economics. The potential contribution to public administration—the institutions and activities of the executive branch—is significant; the marriage of bounded rationality (Simon, 1955) to “predictable irrationality” (Ariely, 2008) can bring new insights to public administration for improved outcomes.

The intersection of behavioral economics and public administration—what Tummers et al. (2016) and Grimmelikhuijsen et al. (2017) have coined “behavioral public administration” (BPA)—is situated in the broader context of a global emphasis on inter-/multidisciplinary approaches to research and practice. This reflects the broad trend of inter-organizational studies that permeates the current public administration research literature. Behavioral economics is at the vanguard of this movement as its application to other social sciences affirms its own relevance as a mix of social psychology and political economics (Kahneman, 2003). The growth of “nudges”—a policy that employs behavioral insights to counter biases in decision-making (Benartzi et al., 2017; John and Stoker, 2017; Oliver, 2015; Thaler and Sunstein, 2008)—from financial interests to health care and environmental issues, attests to its potential for broad social impacts.

Public administration, as an academic discipline and bureaucratic practice, depends on a coherent body of theory to legitimize its authority for telling people what they can and cannot do. Behavioral economics prompts public administration to reorient its perspective from what regulation and policy dictates it to do, and toward a position that initially considers what people are likely to do and then revises policy to accommodate it in an appropriate manner. This is a change from the traditional bureaucratic modus operandi, notwithstanding public choice or the notion of public administration as the execution of the democratic will. To that end, a functional theory of BPA must be developed to situate the research and practice in various policy contexts.

This article looks to fundamental concepts of administration that are complemented by behavioral economics to shore up the foundation for a theory of BPA. The format is a qualitative inquiry that presents: a synopsis of behavioral economics philosophy for public administration; a brief review of selected literature; the recounting of several current BPA examples and practices; prospects for future BPA applications; and the formulation of the theory with propositions. This article takes a high-altitude pass over developments that are contributing to the formation of BPA, noting the higher peaks but not focusing on all features of the landscape. The perspective is primarily looking at decisions and interactions by citizens in the context of administrative policies, rather than including such fields as decision theory, behavioral psychology, or even organization theory, which considers the decisions inside a group. The intended purpose of a theory of BPA is to enable further research and the eventual integration of behavioral insights into the design and implementation of policy for better outcomes.

Behavioral economics in public administration

Taking the position that the purpose of public administration is to provide solutions to public problems, this entails an understanding of how and why those problems exist in order to meet them with policies. For the most part, the problems are met by attempts to impose rational explanations for their existence and rational policies to correct the situation in the traditional economic mold of decision-making. Many day-to-day problems are met adequately in this manner, be they instances of information asymmetry, conflicts of interest, or market failure, as public administration has appropriated lessons from traditional economics in its decision-making processes. Much of the internal working of bureaucracy is based on cost–benefit analysis, which assumes the paradigm of utility-maximizing rationality.

The conflict between the theories of behavioral and standard economics is centered on questions of human motivation in decision-making. Empirical evidence is accruing against the idea that human behavior is adequately explained as maximizing self-interested utility values, much less that we are consistently guided by rational cause–effect calculations (Jones, 2017). That we do not always seek profit or make economically optimal decisions is befuddling to traditional models, especially when regarded from a bureaucratic-rationalist perspective. It presents a philosophical conundrum: how to evaluate behavior absent a priori, infallible, and fundamental tenets for our cognitive operations. The best idea available to accommodate the variety of rationalities is affixing the label “multiculturalism” or “value pluralism” and qualifying situations as being contextually determined, while still harboring rational economic tendencies behind concessions to the tagline: “it depends.”

Behavioral economics seeks to understand instances when people do not act in accordance with principles of utility maximization and make choices that do not cohere with their assumed best interests (Ariely, 2008; Jones, 2017). Before the advent of behavioral economics, bad decisions were dismissed as faulty thinking or misappropriated heuristics, rather than considered as potentially valid alternative choices under different objectives. Simon (1955, 1997) offered “bounded rationality” as an excuse to legitimize the human condition in organizations, yet it sidesteps situations wherein the decision is something other than normatively bad. While heuristics may become outdated and lead to suboptimal outcomes, they nonetheless result from a natural process for efficient problem solving that cannot be dismissed out of hand.

The implications for public administration theory are significant. If the epistemic model for governance—used here to indicate the activity of government—is a standard rationality and yet society displays behaviors that do not follow the standard, then the mismatch must be resolved through some concessions in the epistemic model. There are demonstrated instances where the bureaucratic understanding of how people should behave does not cohere with how people actually do behave, such as expecting people to respond positively to the obvious need for certain disease vaccinations but not getting full compliance by the public. While nobody wants to get sick and it is perfectly rational to get a vaccine, people still neglect to do so for a variety of reasons that are not accounted for in the current administrative episteme.

This seems a nod to postmodernism in that public administration needs to accept alterity and meet the varieties of behaviors on the part of the public with a variety of tools on the part of bureaucracy. This is an opening for theory to develop; the empirical evidence that people behave contrary to administrative expectations demands a reconciliatory response. Research has advanced to this point (Grimmelikhuijsen et al., 2017; James et al., 2017; Jones, 2017) and it is time to formulate a theory in this vein.

Theories in public administration must pay special attention to public service values, such as democratic participation, the rule of law, transparency, accountability, governance ethics, efficiency, effectiveness, and equity. Overlaying these is a broader notion of morality, which is perhaps the greatest source of divergence for public administration from economics, implicating behavioral economics as similarly amoral by association. Sunstein (2016), Oliver (2015) and the Nudge movement (Thaler and Sunstein, 2008) have done yeoman’s work in raising questions about the ethical dimensions of nudges, yet there is still the need for a coherent theory of BPA that will allow the dominant praxes—such as performance management, public–private partnerships, and e-government—to integrate behavioral economics lessons.

Literature review: the neighborhood of the intersection

This research echoes and extends several of the recent works on behavioral administration (Benartzi et al., 2017; Grimmelikhuijsen et al., 2017; Jones, 2017; Oliver, 2015), while also faceting the theoretical angle for how BPA can be applied in governance. The objective of this section is to recount a few important works that go toward the formulation of a theory of BPA without getting swamped in the entire history of behavioral economics. The foundation for this research goes back 60 years to Herbert Simon (1955, 1997), with sporadic contributions over the interim (Arrow, 1974; Bretschneider and Straussman, 1992; Jones, 2003; Lynn, 1986) until the onset of a renewed interest and study of behavioralism in recent times (Hursh and Roma, 2013; McAuley, 2013; Marvel, 2016; Olsen, 2015; Weaver, 2015).

For the purposes here, there is a conveniently prescient provision from Herbert Simon in an operational definition that differentiates the shift from standard economics:

Behavioural economics is concerned with the empirical validity of these neoclassical assumptions about human behavior and, where they prove invalid, with discovering the empirical laws that describe behavior as correctly and accurately as possible. As a second item on its agenda, behavioural economics is concerned with drawing out the implications, for the operation of the economic system and its institutions and for the public policy, of departures of actual behavior from the neoclassical assumptions. A third item on the agenda is to supply empirical evidence about the shape and content of the utility function … so as to strengthen the predictions that can be made about human economic behavior. (Simon, quoted in Eatwell et al., 1987: 221)

The journals have published articles at irregular intervals, suggesting that public administration has been curious about—but not committed to—embracing behavioral economics. On the heels of the first major conference on behavioral economics at the University of Chicago in 1986, Lynn (1986) began an assessment of how policymaking is connected to behavioral sciences. Bretschneider and Straussman (1992) conducted an experiment testing a very common behavioral economics topic—probability judgment—in the governance context, finding that people have a hard time adhering to standards of rationality. Jones (2003) did a more thorough expansion of the relationship between behavioral economics and policy decision-making, echoing Simon’s (1955, 1997) position “to advocate a solid behavioral base for the analysis of political and economic systems” that will result in “a public administration and public policy oriented more toward the dynamics of information processing and away from a fascination with control” (Jones, 2003: 10).

More recently, McAuley contributed a summary of how behavioral economics could apply in policy design, suggesting that, “in its reliance on evidence, [it] can bring more rigour into policy-making, thereby helping regulators make lighter but more effective interventions in the market” (McAuley, 2013: 29). Hursh and Roma (2013: 98) posit that “behavioral economics can serve as a broadly applicable, conceptual, methodological, and analytical framework for the development and evaluation of empirical public policy.” After outlining instances where behavioral economics excels in the explanation of market anomalies, they offer potential applications for several public policy concerns, with the caution that “the variability observed in any behavioral-economic measure must correspond to people’s behaviors and choices in the ‘real world’ that public policy seeks to shape” (Hursh and Roma, 2013: 121). The proposal of BPA by Grimmelikhuijsen et al. (2017) is built on a trend of research in the field that may inspire structured scholarship and academic programs.

Several edited collections tout the virtues of behavioral economics for administrative applications (Schmid, 2004; Shafir, 2013), as well as the best-selling Nudge (Thaler and Sunstein, 2008), which introduced the eponymous term to behavioral economics’ modern vocabulary. A recent “virtual issue” centered on BPA in the Journal of Public Administration Research and Theory (Tummers et al., 2016), a special symposium issue of Public Administration featuring various BPA experimental research, and an increasingly frequent appearance of articles in other public administration journals have driven the point home in the academic literature. The nascent Journal of Behavioral Economics for Policy is emerging as a repository for early experimental research on nudges, and, more recently, the Journal of Behavioral Public Administration has been launched.

Current practices for BPA

The question of how behavioral economics can influence public administration poses the effect it has on affects: does it change the perception and experience of government in a good way? This section presents four major behavioral economics concept groupings for public administration contexts, with examples of successful realizations and extensions of concepts that might yield beneficial outcomes for governance. The concepts described in this section can be—and, in some cases, have been—shaped into testable hypotheses for empirical research. One aspect to keep in mind regarding these concepts is that they are essentially cost-free to integrate, entailing only careful forethought in how policy is presented. Indeed, many of the problems identified by behavioral economics orbit issues of information asymmetry that can be rectified through better education and awareness of conflicts between optimal choices and heuristic patterns.

Future directions in BPA

There are many more concepts for application in governance beyond anchoring, framing, discounting, confirmation bias, choice architecture, and social proof. Indeed, there may yet to be discovered behavioral economics concepts that can emerge from anomalies observed in governance contexts that will contribute to other disciplines. In the meantime, this emerging field is missing a supply of evidence from experiments whereby BPA is tested against traditional administration. Such tests would assess comparative benefits in efficiency, effectiveness, equity, and other public service values. Following the compilation of empirical results and further extensions of behavioral economics concepts, public administration can then use BPA at its discretion, rather than just through special programs blessed by the executive (SBST, 2016). From that point, a theory of BPA can be codified in the framework of administrative philosophy. The theory can feed the further development of behavioral economics concepts that are indigent to governance itself, rather than just modifying the existing concepts into the panoply of administrative practices.

One problem with adapting behavioral economics into the public service context is that experiments testing the effectiveness of concepts have been conducted with college students, game-theory subjects, Mechanical Turk surveys, or other constructs that show how things work in one particular experimental setting, but not necessarily holding external validity for broader application. The irony of behavioral economics is that while it tries to predict irrationality, it then forms a different rationality that may also be unpredictable when tested in alternate contexts.

Furthermore, there is need for cross-cultural examinations of BPA as the preponderance of research has been from developed Western nations’ perspectives. It is incumbent on BPA to demonstrate its validity in a commensurate manner, meeting empirical reliability and standards of rigor for broader generalization. An intriguing synthesis of behavioral sciences and Big Data analysis for policy innovation is proposed by Ruggeri et al. (2017), which essentially turns the traditional method of theory-guided empiricism on its head. While social scientists balk at putting data before theory, the obvious retort is that BPA hinges on instances where the theory and the behavioral data do not fit. This necessitates an alternative—and possibly inverted—approach to empirical policy research.

This runs into two significant issues: (1) “Should a paradigm changing field have to play by the old rules to be acceptable?”; and (2) “Is it really possible to have credible tests of public policy where contextual differences are often so material to the effects as to challenge any hopes for generalizability?” Furthermore, we may add questions about the ethics of employing bureaucratic “tricks” to test theory in the public sphere, especially if public administration blurs lines between influence and compulsion (Sunstein, 2016). Weaver’s (2015) work notes that the use and effectiveness of BPA strategies is highly dynamic, that is to say, that public administrators must be wary of unintended consequences or reversals in effectiveness that may come about from such strategies.

There are promising examples of research that bolster the foundation for more experiments. May (2005) tested hypotheses concerning compliance motivations that are coherent with the kind of carrot-or-stick questions posed in behavioral economics, concluding that traditional regulatory tools are inadequate to meeting administrative objectives. Mann and Wüstemann (2010) argue that behavioral economics helps to resolve information asymmetries if public administration takes an active role. There is also an expanse of research to be conducted on taking advantage of process utility over outcome utility, as evidenced by people’s motivation to vote (Rogers et al., cited in Shafir, 2013: 91–107) and the broad trend for public administration to encourage participatory democratic processes.

Outlining a theory

Based on the current and potential applications of behavioral economics in public administration, it is now possible to outline a theory of BPA with propositions as to what it may empirically accommodate and hypothetically enable. This must be situated on the operative plane of governance: a space described by the dimensions of the people and public administration. The array of behavioral economics concepts are applied to two groups of disconnects: from behavioral economics, there is the difference between what people think they want and what people actually do; and in administrative behavior, there is the gap between what public administration is expected to do and what public administration actually can do. The space in between is where the theory of BPA is located (see Figure 1).

OPEN IN VIEWER

Figure 1.

The location for a theory of behavioral public administration.

BPA meets at the intersection of the two disconnects to address what people expect and what they receive by extending a third dimension with the provision of what people ought to get. Policy has been located somewhere on the plane comprised of ideals, disappointments, and compromises, with only the traditional rationale of economic utility-maximization thinking. The BPA dimension lifts policy possibilities from this flatness to allow greater access to the “ideal” quadrant; we might imagine a curved vector emerging that bends administrative potential in that direction for more desirable outcomes of governance. In essence, BPA offers a reconciliatory approach between citizens and policy, bringing the empirical-positive aspects of their behavior into alignment with the objective (normative) aspects of policy designs. The theory of BPA proposes a corrective measure for the disconnect between “should” and “would.”

There need to be conditions that satisfy certain criteria to justify the use of BPA, such as: evidence of anomalies in rational behavior—that is, the common biases found in behavioral economics—on the part of individuals and institutions; information asymmetries that cause inequitable distributions of benefits or failures of market mechanisms; and instances of unexplained non-compliance with otherwise harmless regulation. As Sunstein (2016) views it, government is justified in using nudges and choice architecture to improve outcomes if it upholds people’s welfare, autonomy, and dignity. Yeung (2016) echoes this position, with an emphasis on individual freedom. This coheres with the objective of the theory offered here, in that public administration should be striving to deliver the ideals of what people ought to get, using tools that can transcend the compromises and shortcomings of rational behavior.

To this end, the following three propositions may provide guidance for the testing and application of the theory, as well as sparking discourse for further refinements to the theory. These are derived from the extant literature, as well as the trends and analysis discussed in this article. The propositions argue for the continued development of BPA by virtue of what has already been demonstrated as effective and what can be possible given the current momentum.

Proposition 1: BPA is an alternative toolset for optimizing the implementation of policy. The curve of the policy vector stemming from the government and citizen intersect is nudged toward the ideal outcomes and away from compromise or failure outcomes if BPA supplements the traditional processes to accommodate non-rational anomalies in citizen or bureaucratic behaviors. This may be verified as research develops to compare the outcomes of BPA policies versus those of the traditional ilk.

Proposition 2: BPA is ethically sound and upholds administrative values. Assuming that BPA adheres to the deontological precepts of democratic governance, such as protecting individual choice and maintaining transparency, wielded from an attitude of light paternalism, then it is coherent with the ethics and values of public administration. Recognizing that there are different perspectives on utility, BPA implicitly legitimizes value pluralism and thus serves the democratic will. As there are continued questions about the ethics of privatizing and other bureaucratic behaviors, it seems necessary to iterate that BPA is actually respecting ethics at its very foundation through transparency and greater information provision.

Proposition 3: BPA research is a unique stream of policy experimentation and methodology. The calls to shift perspective from the importing of behavioral economics tools to the development of behavioral administration tools speaks to the validity of BPA as a unique school in administrative theory (Grimmelikhuijsen et al., 2017). This can inspire the establishment of academic agendas to continue developing BPA in conjunction with practice, as demonstrated by the introduction of dedicated journals and symposia. There is opportunity to focus the behavioral economics lens on the functioning of bureaucracy itself to contribute to governance from within its own institutions, including the investigation of the biases and assumptions of rational paradigms in current administrative thinking.

These propositions pave the way to advance practices in public administration that are not adequately served by current policies. For example, New Public Management’s espousal of “bottom line” policies only accounts for utility-maximization behaviors by both bureaucrats and citizens, rather than taking into consideration the subjective values that influence individual choices away from simple rational efficiency. Basing implementation guidelines on observed behaviors will help integrate the inherent biases that individuals demonstrate, such as the social norm of “bureaucrat bashing” that prompts a knee-jerk resistance to even well-intentioned policy. The ensuing hypotheses for BPA may be better at identifying factors that improve effectiveness and outcomes (in preference to efficiency and outputs), with models that habituate biased behaviors, as well as hypotheses that address big questions and core values of public administration, such as democratic participation, transparency, accountability, and equity. This may result in research with enhanced external validity insofar as there are no presumptions of economic rationality in the range of allowable data. Furthermore, we now have some understanding of alternative stimuli and non-traditional behavioral relationships that exist in the interactions of administration and the polity.

Conclusion

A slew of bestseller books, a series of Nobel Prizes in Economics from associated scholars (Richard Thaler in 2017, Angus Deaton in 2015, Daniel Kahneman in 2008, Gary Becker in 1992, Herbert Simon in 1978, and Kenneth Arrow in 1972), and even a burgeoning industry in behavioral economics policy consulting suggest that the prospects for its use are growing along a demand curve. As the SBST reported modest success (Benartzi et al., 2017; SBST, 2016), we can hope that such work will continue to test BPA in more administrations.

Future research in BPA will need to identify actual changes in policy formulation and implementation practices that can achieve greater compliance with policy objectives. What specific advances have been made in this vein compared to traditional approaches? How do bureaucrats inside agencies and at street level perceive BPA in terms of equity, efficiency, and effectiveness? Is this a fad or a real opportunity for public administration to develop novel means for improving public service? Do the effects of BPA fade over time as populations become acclimated to nudges? These questions must be considered in experiments seeking to extend BPA’s internal validity into the policy arena for tangible, experienced benefits to society, as well as illuminating the strengths and weaknesses of the developing theory.

One last aspect to address is what exactly BPA is: a subset of public administration, a methodological discipline unto itself, or a hybrid of hybrid social sciences that operates above and beyond its own foundations? The reflection by Grimmelikhuijsen et al. (2017) posits BPA as a compliment to traditional public administration. This article advances BPA as a theory that takes a more activist perspective toward governance, envisioning a public administration that is committing to objectives by employing sharper tools made from stronger empirical materials. This analysis and discussion of BPA should enable public administration to recognize the assumptions of behavior in policies and, if necessary, the potential means to make those policies better suited to demonstrated behavior.