Abstract How to account for the ubiquitous presence of public sector performance regimes given evidence that such regimes have failed to achieve their promises? We argue that this paradox perseveres partly because the dominant doctrinal approach—justifying what we label as the external accountability (EA) regime—responds to a real need for political account giving, but also partly because alternative frameworks that can satisfy that need have been slow to emerge. We describe a fundamental mismatch between EA regimes and the basic characteristics of the public sector. As a result, performance regimes are often experienced as externally imposed standards that encourage passivity, gaming, and evasion, and will therefore never be able to achieve performance gains that depend on purposeful professional engagement. Rather than simply criticizing EA regimes, we offer an alternative framework better informed by the empirical study of performance regimes. The proposed alternative internal learning regime makes the case for extensive professional involvement in the development and interpretation of goals. This type of regime offers not just a framework to inform the design of performance regimes, but also a prospective research agenda to address how performance regimes affect motivation, behavior, and public sector outcomes. Introduction The turn to performance measurement practices in the governance of public services is consequential enough that scholars of public organizations have proposed the term “performance regimes” to “capture the embedded nature of these practices in almost all aspects of contemporary governance” (Moynihan et al. 2011: 141). This trend has not only outlasted New Public Management, but also appears to escalate once in place (Pollitt et al. 2010). A doctrine of performance management (Moynihan 2008) was elevated to such a sacrosanct status that its theoretical and empirical foundations have rarely been questioned in public debates (Dubnick 2005; Pollitt et al. 2010). Performance regimes based on external accountability (EA) have, however, created remarkable contradictions in contemporary governance. First, although such regimes are premised on greater managerial autonomy, resistance has risen among public managers and especially frontline staff as they are disempowered and exposed to increased external control (Jakobsen and Mortensen 2016; Soss, Fording, and Schram 2011). Second, such regimes are based on the doctrinal belief that external political accountability increases performance. This, however, contrasts with the growing evidence that EA itself generates large transaction costs that may not be offset by gains in efficiency (Bischoff and Blaeschke 2016; Dubnick 2005) and unintended effects such as goal displacement and gaming. It is far from certain that the net impact of performance management has been positive (Heinrich and Marschke 2010; Pollitt 2013: 358). A recent meta-analysis of public organizations shows only a small positive average effect on performance (Gerrish 2016). This article seeks to make sense of these paradoxes while identifying an agenda for research into performance regimes. We do so by offering a framework grounded in theory and evidence on performance regimes in complex public services. Although administrative doctrines such as the performance management doctrine consists of prescriptions based on claims of cause-effect relationships (Hood and Jackson 1991: 13), our framework consists of a set of assumptions, specified scope conditions, and empirical expectations. To arrive at this framework, we conceptualize performance regimes, divide them into subtypes, and scrutinize them in the context of three stylized facts about the context of public services: Political demand for account giving. Public organizations must account for performance (Dubnick 2005), and performance metrics are politically viewed as a legitimate and necessary way of doing so (Moynihan et al. 2011). Task complexity. Task complexity (Campbell 1988) makes both monitoring and incentivizing more difficult (Dixit 2002) while also making capacity, professional norms, and autonomous motivation more important for performance. Motivation. Autonomous motivation is enhanced by participation in goal setting but crowded out by externally imposed goals and measures (Weibel, Rost, and Osterloh 2010). These stylized facts are clearly broad simplifications but at least have the virtue of being explicit claims about the state of the world (whereas claims in administrative doctrines are often implicit, making them harder to rebut) and hint at scope conditions about how broadly our claims should be applied. We also assume that political principals view regimes primarily as mechanisms to provide quality services to the public that will be rewarded with positive election outcomes. This assumption seems straightforward enough (see e.g., Gailmard & Patty, 2013), but other theories of bureaucratic politics suggest that principals prefer regime control over performance (consistent with Moe, 1989). Even if a preference for control holds, there is still a value in illustrating the costs that such a preference raises. We examine an EA regime, which is the dominant performance regime type of contemporary reforms. In this regime, externally set goals are tied to incentives (Heinrich and Marschke 2010; Pollitt et al. 2010). We contrast the EA regime with a professional regime that provides autonomy in terms of goals and does not link goals to incentives. We offer a hybrid alternative that falls between these two regimes in the form of an internal learning (IL) regime, which combines political-bureaucratic performance goals with extensive professional involvement in the development and interpretation of goals. The currently dominant EA regime aligns with the need for account giving through performance metrics but not with task complexity or autonomous motivation, which are particularly prevalent in welfare policy areas such as health care provision and public education. On the other hand, the IL regime is better aligned with high task complexity and autonomous motivation. By developing types of performance regimes and relating them to the needs of practical public governance, one contribution of this article is to answer the call to public management scholars to address important questions such as how to design effective governance regimes (Kettl 2015). We are also motivated by the conviction that it is not enough for public administration scholarship to simply critique the flaws in administrative doctrines. As long as policymakers need coherent organizing frameworks to guide choices, criticism by itself does not answer the question about what they should do instead. Our approach follows a “design science” premise (Barzelay and Thompson 2010) that scholars ought to apply what we have learned from decades of reforms to inform the future design of public organizations. Although there is a growing and impressive literature on performance information use (Baekgaard and Serritzlew 2016; Kroll 2015), it only touches on broader questions of the design of performance regimes. We aim to incorporate these micro-level insights with the design of macro-level performance regimes as understanding the motives and behavior of individuals working in the public sector is essential to designing good systems (Simon 1947). In the context of the important question of how to design public performance systems, our analysis suggests that the contradictions of performance regimes begin to make sense when interpreted as the result of using EA regimes in situations that are better suited to IL regimes. The mismatch between regime type and a complex public context is in our framework expected to result in disempowered and less motivated employees, weaker performance, and negative unintended effects. Although there is ample scholarly criticism of the EA regime, this has not cohered into a viable alternative. Without such an alternative, the performance management doctrine justifying the EA regime—even if badly frayed—will continue to set the terms for both research and practice (Dubnick 2005; Pollitt et al. 2010). Several scholars (Pollitt 2013: 346; Talbot 2010: 121) have called for research that more deeply interrogates the underlying logics and effects of performance regimes. There is also a real-world demand for such an alternative. For example, in the United States, policymakers are beginning to rethink the strong incentives of the Race to the Top and No Child Left Behind (Mathis and Trujillo 2016). In Denmark, regional governments are experimenting with alternative governance regimes for hospitals, altering or abandoning external financial incentives (Søgaard et al. 2015). Even early proponents of the performance doctrine, such as the World Bank, are searching for new ways of understanding performance management (Moynihan and Beazley 2016). Our approach is not wholly novel; indeed, it builds on previous examinations of long-standing tensions within the public debate on performance management in relation to control and incentives on the one hand and freedom and autonomy on the other (Aucoin 1990: 126; Kettl 1997; Mashaw 1983; Pollitt 2013: 358). Therefore, another contribution of our article is to offer an alternative theoretical framework to EA regimes and the doctrine of performance management that reflects the complexity of public functions. In doing so, we are wary of making excessively broad claims characteristic of administrative doctrines (Hood and Jackson 1991) that, at their worst, consider evidence selectively and do little to establish the scope domain of claims. However, doctrines are consequential (Barzelay 2001; Moynihan 2008), and scholars ought to offer theory- and evidence-based frameworks that counterbalance the less carefully developed doctrines that policymakers turn to. The framework proposed in this article should allow both scholars and practitioners to move beyond the ordinary knowledge and rhetoric typically driving changes in dominant administrative doctrines (Hood and Jackson 1991: 7–11) towards a more scientifically grounded design of performance regimes. The framework builds on and combines many existing strands of research that provide concepts, hypotheses, and empirical knowledge on performance regimes. We do not provide a new theory at odds with these strands of research but instead seek to combine them into a broader theoretical framework. This framework should constitute an alternative to the performance doctrine and outlines research questions and empirical questions that can be settled based on empirical evidence. The article is divided into four main sections. In the first section, we present and analyze EA regimes and the theoretical and empirical critique of such regimes. Next, we discuss traditional alternatives to performance regimes, followed by the conceptualization of the IL regime. Throughout the article, regimes are analyzed within the domain of complex public tasks. We end by discussing how our analysis provides insights into the paradoxes of performance regimes while offering avenues for future empirical research. The EA Regime and its Challenges Performance regimes consist of specific combinations of external actors with the opportunity to steer the performance of public organizations and the specific means by which they do so (Talbot 2010). In our conceptualization of the EA regime, these external actors are political and bureaucratic superiors that use the means of external goal setting and incentives. More specifically, the EA regime has two defining features (Boyne 2012, 209; Heckman, Heinrich, and Smith 2011, 1): Goals are externally determined by political-bureaucratic actors and operationalized in performance measures to track activities. High-powered external incentives connect the achievement of goals to rewards and punishment, primarily through budget allocations. Drawing—even if loosely—on principal-agent theory, the core assumption of the performance management doctrine is that the EA regime increases performance by imposing goals, collecting information on goal fulfillment and tying incentives to goal fulfillment. Hence, the EA regime encourages organizations to respond to the goals of their principals. As such, the EA regime directs attention to specific goals and allows political and bureaucratic actors to draw out hidden information. For example, in the field of health care, they allow politicians and managers to observe the opaque medical profession by showing the differences in costs of specific treatments between care providers. Consistent with these claims, some studies have shown performance improvements after implementing an EA regime with externally set goals connected to incentives (Bevan and Wilson 2013; Kelman and Friedman 2009). The doctrine of performance management also implies a trade of managerial autonomy in return for external influence over performance goals and measures (Behn 2002; Moynihan 2008). Managers receive more autonomy to manage, whereas in return, the political and bureaucratic superiors hold them accountable for performance. EA regimes, however, face some important challenges in relation to the other two assumptions regarding the delivery of complex public tasks: task complexity and autonomous motivation. Task Complexity The complexity of a task increases with the amount of information that a rational decision-maker must take into consideration when solving the task (Campbell 1988; Latham and Yukl 1975; March, Simon, and Guetzkow 1993). In complex public services such as health care provision, this includes the following: Multiple performance dimensions (Boyne 2002) that generate multiple goals, such as the quality of service, accessibility, efficiency, equity, and patient satisfaction Multiple paths to reach the goals through different treatments and services Tradeoffs between goals (and paths) such as between quality and accessibility Effect uncertainty regarding how different choices in, for instance, health treatments will affect specific patients as the capacity and needs of patients even with similar diagnoses can differ substantially. Many complex services are hence delivered by professionals that have had years of training and experience building (tacit) knowledge on how to handle each particular case (Freidson 2001; Noordegraaf 2007). This contrasts with simpler services such as garbage collection, where the goals—trash collected—and means to the collection—garbage trucks making rounds—are simpler and easier to comprehend, plan, and observe. Although the EA regime is well suited for simpler tasks, many public services have some level of task complexity that challenges the EA regime and its supportive doctrine of performance management. The overall challenge of high complexity to the EA regime relates to the key doctrinal assumption that “measures capture performance”, which states that it is possible to generate measures that form valid and comprehensive indicators of performance (Hood 2006). With complex tasks, this is highly unlikely. Instead, many complex tasks in public service can only be measured through incomplete and imperfect measures that only cover some of the organization’s performance goals. This challenge helps to better understand the paradoxes faced by EA regimes. First, when only some dimensions of performance are measured and rewarded, organizational actors will have an incentive to direct attention to the aspects of performance to which they are held accountable and neglect unmeasured aspects (Blau and Meyer 1971; Dixit 2002; Holmstrom and Milgrom 1991). This will lead to goal displacement (Merton 1940). Indeed, there is substantial empirical evidence of displacement towards measured versus unmeasured aspects of performance (Ordóñez et al. 2009). Kelman and Friedman (2009) show that an EA-like regime improved not only emergency waiting times but also the general quality of care. However, this improvement may have been made at a cost of other aspects (Pollitt 2000), such as investments in training of employees or preventive care. Second, if the measure of a particular performance goal only represents one aspect of the goal, it creates incentives for undesired selection effects (Bevan and Hood 2006). For example, organizations may focus attention only on the subset of clients most likely to lead to reward and ignore or systematically exclude those unlikely to generate a reward. Practical examples include schools focusing only on students near the passing grade for a test while neglecting others (Figlio and Rouse 2006), job trainers failing to provide adequate help to candidates deemed unlikely to receive offers (Heinrich and Marschke 2010; Koning and Heinrich 2013; Soss, Fording, and Schram 2011), and hospitals creaming profitable or dumping unprofitable patients (Cots et al. 2011). Third, performance data is likely to be subject to gaming and manipulation by the agents that it is intended to hold accountable. In the area of education policy, for example, the link between performance measures and rewards has incentivized cheating by teachers on behalf of their students (Jacob and Levitt 2003). A particular form of gaming is to manipulate the measurement system via strategic reclassifications. Schools might reclassify a higher proportion of students as disabled in order to exclude them from testing evaluations (Figlio and Getzler 2006). Health providers may systematically reclassify patients as belonging to a category of services that provide higher reimbursement rates (Kerpershoek, Groenleer, and de Bruijn 2016). Gaming and manipulation can occur when agents have some control over the collection of performance data or where agents can manipulate work processes to give a misleading impression of success. Manipulation and gaming is hence a particular challenge for complex tasks. For example, since doctors register their own diagnoses and treatment outcomes, there is a high reliance on their capability and willingness to provide valid information. When agents do not identify with a regime—which is likely with externally imposed regimes—there is a larger risk that employees will not implement it well (Tummers and Van de Walle 2012). The corruption of data also becomes more likely when the incentives to perform are strong. Campbell’s Law, articulated by Donald Campbell (1979: 85), notes how measures collected for social science research corrupts when used for allocation of resources: “The more any quantitative social indicator is used for social decision-making, the more subject it will be to corruption pressures and the more apt it will be to distort and corrupt the social processes it is intended to monitor.” Hence, measures connected to external incentives are more likely to be subject to gaming and manipulation. Some incentives may have more severe impacts than others. For example, Bevan and Wilson (2013) show that naming and shaming may be beneficial for performance. Although this is an external incentive, it is a milder form than financial incentives that directly threaten the continuation of the service provision. Whereas financial incentives speak to the extrinsic motivations, naming and shaming incentivizes the internal status and ego motives through introjection (Deci and Ryan 2000). Fourth, even when no goal displacement, selection effects, and gaming occurs, external performance-based systems often fail to match the dynamic nature of complex services (Heinrich and Marschke 2010). As it is impossible to capture all possibilities in rules and guidelines, the EA regime can lag behind in reality. The above challenges are not particular for public sector organizations but are a general problem for organizations facing complex tasks. As Lazear and Gibbs point out in their introduction to personnel economics (Lazear and Gibbs 2015), the use of strong incentives becomes problematic when tasks are interdependent, there are multiple goals and tasks, agents have little control of the outcome, and there are ample opportunities for gaming. A recent example of misaligned EA regimes in the private sector was the failings of the Wells Fargo scandal. Here, employees met highly incentivized performance goals by providing unrequested services for which they charged unknowing customers. The massive scale and routine nature of the fraud pointed to a regime failure rather than the misdeeds of a handful of “bad apples” (The New Yorker, September 12, 2016). Employee Motivation The performance management doctrine underlying the EA regime makes assumptions regarding employee motivation. Motivation is a combination of energy and direction, and self-determination theory points to the importance of whether motivation is determined autonomously and arises from within (Deci and Ryan 2000). The more controlled or extrinsic the motivation, the more factors external to the person, such as salary and status, drive behavior. Research on public service providers has shown that autonomous motives play an important role for employees. In particular, intrinsic motivation—emanating from enjoyment in doing the task in itself—and public service motivation—identifying with and feeling an internal obligation to contribute to society—seem to drive employee behavior, and especially for highly professionalized employees working on public tasks (Moynihan 2010; Perry, Mesch, and Paarlberg 2006; Perry and Wise 1990). Historically, complex services such as education and health care have benefitted from service providers deriving motivation from the very act of performing the task or from status and honor, professional norms, and the altruistic benefits of serving others. Those with a higher public service ethos are also more likely to enter into public work, less likely to exit, and more committed when in public employment (Perry, Mesch, and Paarlberg 2006). By contrast, EA regimes are built on the doctrinal belief that agents have a controlled motivation, which means that they can achieve externally imposed goals via extrinsic rewards and sanctions. The mismatch between the assumptions of EA regimes and evidence on motivation in public settings creates a number of problems. First, several aspects of EA regimes may be a poor fit for highly autonomously motivated individuals. The EA regime is likely to foster goal ambiguity between individual preferences and externally imposed goals (Pandey and Wright 2006), especially if productivity is prioritized over other goals. This has been observed in different public settings but primarily in health, education, and welfare (Brodkin 2011; Dias and Maynard-Moody 2007; Heckman, Heinrich, and Smith 2011; Langbein 2010; Ryan and Weinstein 2009; Soss, Fording, and Schram 2011; Weibel, Rost, and Osterloh 2010). Second, when the EA regime applies extrinsic incentives to push performance, it creates risks of “crowding out” intrinsic motivation (Deci, Koestner, and Ryan 1999; Gneezy and Rustichini 2000) and, in some cases, also public service motivation (Jacobsen, Hvitved, and Andersen 2014). A classic example of this mechanism is teachers who feel pressured to pursue an extrinsically imposed goal of improving test scores at the expense of the goal of improving more general student outcomes (Ryan and Weinstein 2009). The teachers become more cynical and less autonomously motivated in their jobs, scaling back unmeasured effort and focusing only on teaching to the test. Crowding out processes may also occur in the selection of public employees: More intrinsically motivated individuals may exit EA regimes or choose not to join them in the first place (Moynihan 2010). Third, the use of an EA regime for complex tasks will increase the magnitude of the negative impact of the regime on autonomous motivation. With simpler tasks where measures closely mirror tasks, crowding out is less of a concern since rewards and sanctions attached to measures should still motivate greater effort (the so called “price” effect). However, when complexity makes it difficult to truly observe agent effort, there will be a performance loss on unobserved aspects of performance to the point that the crowding out effect outweighs the price effect (Weibel, Rost, and Osterloh 2010). Further, as intrinsic and altruistic motivations are eroded, so too are normative safeguards against gaming and manipulation (Moynihan 2010). Articulating an Alternative Performance Regime The popularity of EA regimes depends partly on its responsiveness to our first stylized fact—the need for account giving—even though it is ill-suited to satisfy the second and third stylized facts—the task complexity and motivational factors present in the public sector. The empirical and theoretical evidence presented above aligns with this simple observation. The problems with the EA regime in public settings with complex tasks propel a search for alternatives. One is to establish more dynamic EA regimes based on the assumption that the principal can modify and adjust the mixture of incentives and monitoring efforts to respond to problems arising from complexity (Cots et al. 2011; Heinrich and Marschke 2010). For example, Courty and Marschke (2003, 2007) describe the adaptations of federal principals overseeing manipulation of job training programs by state governments. The “dynamic principal” perspective recognizes the problems created by the EA regime but responds to these problems by developing better measures instead of addressing the challenges of complexity and motivation. Put another way, this approach assumes that principals and agents have an adversarial relationship and that accountability based on external goals and incentives are necessary to manage agents. It assumes that it is possible for principals to make more precise measures of performance than today and to recognize and remove gaming opportunities. With enough diligence on the part of principals, performance regimes can be improved and solve current problems with gaming (Heinrich and Marschke 2010). However, such an approach still assumes extrinsically motivated actors and that it is possible to solve the problems with external measurement of highly complex services. Moreover, it assumes a limit on the ingenuity of the agent, even though the agent often possesses as much information (or more) as the principal that can be used to their advantage. A highly dynamic regime that adds and modifies goals loses value in prioritizing goals and directing behavior, which is one of the proposed gains of the EA regime. More dynamic performance principals would also require an investment in public sector administrative capacity and resources, which have so far been lacking in the application of the EA regime (Soss, Fording, and Schram 2011), and a determined commitment on the part of the principal to invest their time in management processes. Although the dynamic principal model offers one possibility for dealing with the problems of EA regimes, we ask whether there is an alternative that better accords with our stylized facts about task complexity and motivation. In the next section, we consider in detail the possible alternatives among traditional models of governance regimes. The Building Blocks of an Alternative On the theoretical level, we treat performance regimes as ideal types. Ideal types are theoretical constructs that describe how specific actors, institutions, and behaviors relate to each other in an internally consistent way based on a given rationale (Stout 2010). The EA regime and the doctrine of performance management make up an ideal type as they prescribe a specific institutional setting featuring motivational and informational assumptions that predict a positive performance effect. One possible alternative would be a professional regime. The core idea behind the professional regime is that organizations solving highly complex tasks are best governed through the judgment and expertise of professionals held together by common norms and values (Freidson 2001; Mintzberg 1996; Ouchi 1980). Professional norms and the influence of professions are hence expected to shape work practices (Exworthy and Halford 1999; Mashaw 1983; Noordegraaf 2007). Professionals have preferences about what good service is (Evetts 2003; Roberts and Dietrich 1999) as well as professional norms about appropriate behavior and service quality that counterbalance a myopic focus on economic considerations (Freidson 2001: 127). Our focus on the professional regime differs from the traditional contrast in the PA literature between EA and the bureaucratic regime, where governance is based on input measures such as rules, orders, and budgets (Hood and Jackson 1991; Pollitt and Bouckaert 2011; Weber 1960). The bureaucratic regime is not a plausible alternative to the EA regime as it does not meet contemporary political demand for account giving using metrics (Moynihan and Soss 2014). Moreover, because of its reliance on rules and regulations as a steering mechanism, it is not aligned with highly complex tasks requiring case-specific judgment, knowledge, and discretion as rules are unable to capture fluid situations. Put another way, it is not well aligned with goal-based learning (Moynihan 2005). To conceptualize a hybrid regime, we therefore draw on the EA and professional regimes as ideal types to identify and build on their rationales. Table 1 identifies the important differences between the EA regime and the professional regime. In terms of the substantive autonomy to determine performance goals and measures, the professional regime allows for internally determined goals and measures tied to professional norms, whereas goals and measures are externally determined in the EA regime. The professional regime hence gives professionals a high degree of autonomy in determining how to do their work. The EA regime does imply some managerial autonomy, but in reality, the loss of substantive autonomy highly influences work practices as well. For instance, teachers have to aim for higher test scores and thus have little substantive autonomy in deciding on their outcome measure, but this goal also leads them to focus their teaching on the aspects that the test measures. Table 1. Types of Regimes EA Regime IL Regime Professional Regime Performance regime Yes Yes No Goal autonomy (substantive) Low. External, tied to political-bureaucratic goals Medium. Lower-level goals and performance measures set by professionals High. Goals are set by professionals, tied to professional values Incentives External, high-powered Internal, aimed at norms, values, and peer status Internal, low-powered Accountability External with performance focus Internal accountability embedded within an external regulative bargain and with regular dialogue between principals and agents Internal, tied to professional norms (with a regulative bargain) Rationale for improvements Improvements through external control and incentives Improvement through learning spurred by autonomous motivation Improvement through new professional expertise Controlled motivation; learning about goal achievement but not goals (single-loop learning) Learning on (lower-level) goals and goal achievement infused with values and experiences of experts (single- and double-loop learning) Autonomous motivation; learning on goals and goal achievement through professional norms and goals (single- and double-loop learning) EA Regime IL Regime Professional Regime Performance regime Yes Yes No Goal autonomy (substantive) Low. External, tied to political-bureaucratic goals Medium. Lower-level goals and performance measures set by professionals High. Goals are set by professionals, tied to professional values Incentives External, high-powered Internal, aimed at norms, values, and peer status Internal, low-powered Accountability External with performance focus Internal accountability embedded within an external regulative bargain and with regular dialogue between principals and agents Internal, tied to professional norms (with a regulative bargain) Rationale for improvements Improvements through external control and incentives Improvement through learning spurred by autonomous motivation Improvement through new professional expertise Controlled motivation; learning about goal achievement but not goals (single-loop learning) Learning on (lower-level) goals and goal achievement infused with values and experiences of experts (single- and double-loop learning) Autonomous motivation; learning on goals and goal achievement through professional norms and goals (single- and double-loop learning) View Large Table 1. Types of Regimes EA Regime IL Regime Professional Regime Performance regime Yes Yes No Goal autonomy (substantive) Low. External, tied to political-bureaucratic goals Medium. Lower-level goals and performance measures set by professionals High. Goals are set by professionals, tied to professional values Incentives External, high-powered Internal, aimed at norms, values, and peer status Internal, low-powered Accountability External with performance focus Internal accountability embedded within an external regulative bargain and with regular dialogue between principals and agents Internal, tied to professional norms (with a regulative bargain) Rationale for improvements Improvements through external control and incentives Improvement through learning spurred by autonomous motivation Improvement through new professional expertise Controlled motivation; learning about goal achievement but not goals (single-loop learning) Learning on (lower-level) goals and goal achievement infused with values and experiences of experts (single- and double-loop learning) Autonomous motivation; learning on goals and goal achievement through professional norms and goals (single- and double-loop learning) EA Regime IL Regime Professional Regime Performance regime Yes Yes No Goal autonomy (substantive) Low. External, tied to political-bureaucratic goals Medium. Lower-level goals and performance measures set by professionals High. Goals are set by professionals, tied to professional values Incentives External, high-powered Internal, aimed at norms, values, and peer status Internal, low-powered Accountability External with performance focus Internal accountability embedded within an external regulative bargain and with regular dialogue between principals and agents Internal, tied to professional norms (with a regulative bargain) Rationale for improvements Improvements through external control and incentives Improvement through learning spurred by autonomous motivation Improvement through new professional expertise Controlled motivation; learning about goal achievement but not goals (single-loop learning) Learning on (lower-level) goals and goal achievement infused with values and experiences of experts (single- and double-loop learning) Autonomous motivation; learning on goals and goal achievement through professional norms and goals (single- and double-loop learning) View Large In relation to incentives and accountability, the professional regime has no strong external incentives. This clearly contrasts with the strong externally set incentives of the EA regime that give principals considerable influence over an organization and make its professionals accountable to them (Pollitt et al. 2010). Examples are lists of good and bad schools or publicly displaying the mortality rates of hospitals. Instead, accountability in the professional regime is achieved through internal fora such as professional associations and peer reviews, with little control from external agencies (Romzek and Dubnick 1987). Professionals in such a regime are, however, indirectly accountable to political, societal, and client interests as the autonomy of the professionals is based on an implicit contract according to which they promise to uphold those interests in exchange for self-determination (Noordegraaf 2007). Still, the professional regime is often critiqued for its lack of direct public accountability (Noordegraaf 2007). Consequently, the rationales or doctrinal beliefs for how organizations improve outcomes within the two regimes also differ. The EA regime is based on the doctrinal assumption that improvement occurs when external political-bureaucratic actors set the right incentives for organizations. The organizations will then increase their effort and seek to adjust their practices to achieve a higher level of goal achievement. Organizational learning—understood as the process of “encoding inferences from history into routines that guide behavior” (Levitt and March 1988, p. 320; March, 2010)—will have a single-loop character where inferences about practices but not underlying assumptions about the goals pursued are taken into consideration (Argyris 1980). Nevertheless, if employees fail to identify with externally imposed goals, securing the motivation needed to facilitate genuine learning becomes less likely. It is, instead, more likely that the organization will passively collect information for account giving (Moynihan 2005). Hence, controlled motivation drives the learning process. In the professional regime, improvement originates from learning through professional norms and ideals, and improvements in goal fulfillment are likely to be based on the acquisition of new professional expertise (Freidson 2001). Mintzberg (1996) proposes that for client-based complex services, learning depends on the professionalism and dedication of professional staff. The organizational professionals are, to a great extent, autonomously motivated, and they respond to incentives aimed at their norms and status. Professional values and motives therefore also play an important role in the IL regime. The IL Regime Neither the EA nor the professional regime fully addresses all three stylized facts about complex public tasks, and nor does either a classic bureaucratic or a dynamic performance regime. What sort of regime, then, can accommodate the three stylized facts? To arrive at a potential alternative ideal type to the EA regime ideal type, we return to the stylized fact that the political environment does not allow public services to ignore the demand for performance-based account giving. Such accounts are usually legally required, and many professions and services have endorsed performance measures as a governing principle themselves, even if objecting to specific metrics or practices (Bouckaert and Halligan 2007; Moynihan 2008; Talbot 2010: 1). Hence, we are not suggesting that the EA regime should be fully abandoned. Such a suggestion is not politically feasible, and it ignores the potential for performance measures to help public organizations. Is a hybrid that contains some features of the EA regime but also elements of the professional regime able to overcome the problems of the two ideal type regimes? This section sketches what such a hybrid regime of EA and professional regimes would look like. We do not claim that such a hybrid regime is new from an empirical point of view. Indeed, the actual practice of performance regimes is unlikely to ever fit the ideal types that we sketch out but, instead, actively blends these approaches.1 However, there is not a strong theoretical logic for how these hybrid forms function in the same way that there is a well-developed logic for EA regimes. As a result, the application of a hybrid regime may be misconstrued as a failure to fully implement an EA regime rather than as a managerial approach that is still goal-oriented in purpose. In practice, legal limits and political opposition may dampen the use of extrinsic incentives, or political principals may recognize the value of professional input in goal selection or the diagnosis of problems. An example of such a hybrid would be from the field of job training, in which the US Department of Labor moved away from an EA regime to focus on a new approach for goal setting that was explicitly intended to give state and local partners more control in setting goals and selecting measures and, by doing so, align beliefs and generate a sense of shared accountability (Heinrich 2007). Another example is seen in the application of “stat” techniques to facilitate organizational learning in public organizations, which is popular in the United States and was adopted by the GPRA Modernization Act of 2010 (Moynihan and Kroll 2016). Here, professional input is engaged less in the setting of goals and more in the interpretation of measures to identify program problems and solutions. The approach reflects the limit of political principals—they need professionals to make sense of data—and their leverage—they can exert accountability through social routines that establish norms of accountability rather than financial incentives. The “stat” tool also evolved as it migrated across professions. It originated in policing, where a paramilitary culture allowed political principals to aggressively demand improvement in ways that bordered on humiliation—such as yelling at police officers in front of their peers—and could not succeed in professions that lacked deep hierarchical norms of obedience (Behn 2014; Hatry and Davies 2011). Even within policing, the unrelenting demands for continuous improvement over time was charged with eroding professional norms to the point that other policing goals were damaged, and officers made their targets by hassling citizens or by manipulating performance metrics (Eterno and Silverman 2010). By contrast, within the US federal government, such routines were associated with a more purposeful use of performance data when they featured positive reinforcement (Moynihan and Kroll 2016). A non-US example comes from the effort by the Central Denmark Hospital Region to move away from an EA regime towards a regime in which hospital wards had some financial incentives removed while they were encouraged to develop their own performance indicators (Søgaard, Kristensen, and Bech 2015; Sørensen and Burau 2016). The regional principals perceived the previous model to produce unintended consequences in the form of gaming and selection whereas hampering quality development focused on the creation of value for the patient. Indeed, this example motivated our current effort as political principals sought scholarly insight to do something different but lacked a theoretical model as to how to make such changes. We return to this example in the discussion of a research agenda. Although hybrid regimes exist in practice, and while policymakers may wish to establish such regimes, we see a lack of theory and language to describe, identify, and understand how such regimes work, reflecting a failure to conceptualize a regime that combines professional and performance features. Although policymakers may already be pursuing a hybrid regime in practice, they may be doing so inadvertently, working at the intersection of countervailing political and professional pressures rather than consciously considering how to combine types of regimes. Similarly, although scholars have forwarded impressive critiques of EA regimes, they lack a conceptual language for understanding and exploring a hybrid regime. Describing the IL regime provides a theoretical framework that—to an admittedly limited degree—fulfills this need. The above examples also show the tension between externally imposed goals and professional autonomy. This is a consistent theme across the critiques of the EA regime. Sociologists argue that organizations are compelled to respond perversely to maintain external legitimacy (Sauder and Espeland 2009); psychologists suggest that externally imposed goals are harmful because they reduce autonomy and self-determination (Frey 2000; Frey and Osterloh 2010); economists suggest that such goals allow a self-interested actor to engage in moral hazard to maximize a financial gain (Dixit 2002); and political scientists note that such systems undercut traditions of loyalty between principals and agents (Hood 2011).2 If one accepts the external locus to be the central problem, the obvious alternative becomes to give organizational actors more autonomy in determining the goals and measures of performance. This would also unleash the benefits of professionalism by removing what professionals perceive to be illegitimate constraining standards that weaken morale and encourage gaming (Kerpershoek, Groenleer, and de Bruijn 2016). The regime would be applicable for complex tasks requiring professionalism while reducing the risk of crowding out autonomous motivation. We call such a hybrid regime an IL regime. Table 1 presents the core elements of the IL regime. The IL regime gives organizations and their professionals more goal autonomy in selecting goals and measures as a matter of deliberate design, thereby giving up complete political-bureaucratic control over goals. The above examples all, to varying degrees, feature political principals recognizing that professional expertise is essential if outcomes are to be achieved. The most careful critiques of principal-agent theories in public settings make a similar claim, arguing that such autonomy is crucial if employees are to be motivated (Brehm and Gates 1999; Gailmard and Patty 2007). Further, this expertise cannot be engaged unless professional actors play a role in establishing and interpreting performance measures, which are de facto provisions of autonomy and influence to professionals that will motivate greater effort. The choice of measures should reflect democratic concerns as well as appropriate professional norms, treatments, and techniques. The more specific and the closer to professional practices goals are, the more the professionals, and not outside political-bureaucratic actors, would be the ones to design the measures. Furthermore, the regime promotes learning and capacity building based on performance information because the regime provides the professionals with discretion to pursue some of their own policy preferences, making it worthwhile for them to learn and invest in expertise (Gailmard and Patty 2013). Using a field experiment, Andersen and Moynihan (2016) offer evidence for this claim, showing that school principals who were granted more discretion in hiring invested more effort to acquire expertise in the form of performance data and that these investments were even stronger when the performance information reflected goals that were consistent with the school principal’s autonomous motivation. Such arguments for discretion also underlie the management approach of many knowledge intensive private organizations such as Google’s 70-20-10 rule, according to which employees are to spend 20% of their working time on ideas and projects initiated without external instruction (Garvin 2013; Steiber 2014). The IL regime also removes a tight link between high-powered incentives and performance measures. In this regime, emphasis is on performance information as a learning tool and less on performance information used to create external incentives to motivate (Greve 2003, 2008). Weakening external incentives reduce the inclination to game in order to satisfy external goals. Goal-oriented learning and innovation will become more likely when organizational actors perceive that they work in a setting where acknowledging errors and problems will lead to a dialogue about problem solving rather than punishment. Expanding goal autonomy also creates room for double-loop learning (Argyris 1980; Moynihan 2008). It is assumed that individuals have a mixed set of motives and are, to some degree, autonomously motivated to perform and learn (Deci, Koestner, and Ryan 1999). Thus, individuals want to learn but need the right structures and support to do so. Thus, when professionals identify more with organizational goals because they can play a role in selecting them, IL from performance data that captures those goals is likely to take place. Positive patterns of performance-based learning tend to emerge when organizational routines to learn are established, accepted, and incorporated as professional expertise (Moynihan and Kroll 2016). The examples cited above feature dialogue-centered social routines as a primary basis of engaging professionals, sometimes with and sometimes without incentives. The metaphor of dialogue recognizes not just that professional expertise is essential to learning, but that such learning arises from an iterative engagement between political principals pursuing democratic goals and professional actors providing expertise. Moynihan (2008) identifies the interdependence of dialogue, learning, and professional identity by contrasting three corrections departments implementing performance requirements. In one, the performance regime was seen as externally imposed by political principals, setting unrealistic performance expectations at odds with limited resources, resulting in a passive professional response. In another, the performance regime generated single-loop learning under conditions in which political principals supported goals also valued by corrections professionals. In the third corrections agency, double loop occurred where corrections professionals engaged in a multi-year dialogue about the purpose of a corrections agency. This learning was spurred internally by a clash between professionals with different backgrounds—some with a social science training that challenged those coming from a corrections background. The mix of professional beliefs generated conflict but eventually established a new set of organizational assumptions, which the department put into practice only through a long external dialogue with political stakeholders about the wisdom of the new approach. The case also illustrates a deeper point about the nature of accountability for complex tasks in performance regimes. Such accountability cannot realistically be expected to arise from political principals selecting and monitoring data that they struggle to process. Instead, it depends on a dialogue between principals and the professionals who are best placed to interpret and make sense of ambiguous performance information, and such a dialogue will be most fruitful if it is collaborative. This claim is evident also in the importance that the field of evaluation places on professional knowledge in understanding and interpreting outcomes (Funnell and Rogers 2011, Mayne 2004). It is also present in aspects of organizational learning that emphasize the ambiguity of organizational settings (March 2010). Building on elements of autonomy and the removal of external incentives from the professional regime raises the issue of performance-based accountability in the IL regime. Compared to the EA regime, less direct control is given to principals. One implication is that different organizations may become less comparable when there is no control from principals to harmonize their measures. Our hybrid model still assumes that performance data exists and is publicly reported. Still, it would be naïve to assume constant alignment of the goals of employees and external stakeholders. An obvious point of conflict, for example, is the trade-off between cost containment, quality, and rent-seeking on the part of the profession. There is thus valid concern regarding the accountability of public service in an IL regime. However, the reduced political control under an IL regime does not necessarily mean an overall loss of accountability, given the potential tradeoff between principal control and goal achievement (Gailmard and Patty 2007). Professional norms and values also offer a form of accountability since they are established to demonstrate the legitimacy and efficacy of a profession to the broader public and should thus share common ground with democratic goals to provide public value. Indeed, performance-based accountability should be at least partially ensured by embedding goals within an informal (or formal) regulative bargain (Kerpershoek, Groenleer, and de Bruijn 2016) between the professions and the state and society. The professions are granted autonomy in exchange for regulating their own members in terms of professional conduct as well as broader, set societal goals such as cost containment (Suddaby, Cooper, and Greenwood 2007). Such bargains are struck within a political-bureaucratic structure that allows political principals to fundamentally change the regulative bargain should the professions not live up to their end of it. Any use of substantive autonomy by the professions would thus have to take this shadow of hierarchy into account (Scharpf 1994). Regulative bargains provide a formidable constraint that prevents the adoption of pure EA regimes, especially when professions have strong political standings. Our model adds to the existing consideration of regulative bargains by proposing that they can achieve better outcomes through ongoing routines of goal-based dialogue between political principals and professionals rather than through abrupt disruptions. A New Empirical Research Agenda We have outlined a framework with a potential alternative performance regime better aligned with the conditions of complex public services. Making a design science perspective compatible with social science means not just sketching an alternative regime, but also outlining a research agenda that can determine how the new framework is working and whether other alternative designs would work better. In this final section, we outline how our discussion of performance regimes motivates a new empirical research agenda that can make sense of contradictions in contemporary governance. A key element in the public administration research agenda on performance management research should be to examine the empirical consequences of moving from an EA to an IL regime or from a professional to an IL regime. In the following, we focus on organizations with an EA regime as a starting point because recent decades have seen public organizations move to this regime type, and as cited in our introduction, there appears to be some appetite among policymakers to move in the other direction. In situations in which there has been no such movement to EA regimes, our proposed approach may not be valid, but instead, an appropriate balance between the EA and IL regime implied by figure 1 should be pursued. Even though our goal is a broad reconsideration of the design of performance regimes, such empirical testing requires that our framework offers clear and falsifiable expectations. We therefore offer several expectations about the effects of going from an EA to an IL regime through the removal of external incentives and increased goal autonomy. These expectations are based on the argument that the drawbacks of the pure EA regime and the professional regime can be addressed by an IL regime that gives agents some substantive autonomy in determining what performance is and how it should be measured, that lies within a regulatory bargain, and that removes any tight links between high-powered incentives and performance measures. Figure 1 illustrates the presumed attitudinal, behavioral, and performance-related consequences of moving from an EA to an IL regime. However, not one particular performance regime is best in all contexts. How regime changes affect performance depends on two critical scope conditions. Following the argument of this article, when there is high task complexity and high levels of autonomous motivation, moving from an EA to an IL regime is expected to have a stronger and more positive impact than in cases with little task complexity and low levels of autonomous motivation. Comparing performance regimes in different contexts can test these critical assumptions. For functions that are simple and easy to measure, where it is easy to assign responsibility and minimize the need for discretion, and where the context is stable, something close to ideal-type EA regimes would work the best. Moreover, whether performance regimes influence behavior also depends on the implementation of the regime as the effects may be limited if organizations and employees resist the implementation (Tummers and Van de Walle 2012). Although a shift in regime is likely to affect several attitudes, behaviors, and outcomes, we focus here on key factors that a regime change is likely to influence. Figure 1. View largeDownload slide Model of the impact of performance regime change. Figure 1. View largeDownload slide Model of the impact of performance regime change. In relation to employee attitudes, we expect that internal regimes will increase autonomous motivation and reduce perceptions of red tape. As discussed above, studies have shown that aspects of EA regimes are likely to crowd out intrinsic and altruistic motivation (Gagné et al. 2010; Wright and Pandey 2010). Reducing the external locus of control associated with EA regimes should reverse the negative effects on employee attitudes. The rise in substantive autonomy—having the agent formulate goals and set targets that reflect professional norms—is likely to increase agent identification with the goals (Deci, Koestner, and Ryan 1999), which can be expected to improve autonomous motivation. Finally, the considerable transaction costs that arise with an externally imposed performance regime may be experienced as red tape that employees have to deal with. By contrast, including employees in deciding on the indicators and data collection is likely to decrease these perceptions even if this does not reduce the actual work-load, since employees will better appreciate the instrumental values the performance regime serves (Laverty, Lewis, and Moynihan 2013; Van Loon et al. 2016). While we focus on employee attitudes that are hypothesized as connecting with performance, actual transaction costs are also worth exploring. Proposition 1: A shift from an external accountability regime to an internal learning performance regime will increase autonomous motivation among employees and decrease perceptions of red tape. We also expect changes in intended and unintended output behaviors. A move towards an IL regime will remove perverse incentives to goal displacement, gaming, and other forms of unintended behaviors (Kerpershoek et al. 2016). Furthermore, the shift towards the IL regime removes the disincentive for cooperation inherent in the EA regime, in which rewards and punishments typically relate to the goals of individual organizations irrespective of whether there are larger collective goals requiring cooperation between the organizations. For example, Soss, Fording, and Schram (2011) note that providers of job training services under EA failed to coordinate with one another as incentives encouraged competition rather than cooperation. Under the IL regime, actors are not rewarded for specific outputs provided and therefore do not perceive time spent in routines of dialogue as distractions from the work they are rewarded for. Another example from the field of health care is Gawande’s (2009) attempt to explain variation in costs for health services in different communities. He pointed out that hospitals that employed a mixture of fixed salary (thereby limiting extrinsic incentives) and appeal to professional values through a patient-centered culture were associated with more meetings in which specialists collectively decided on a treatment plan for the patient, which resulted in both better outcomes and lower costs than average. Here, professionals saw a team-based dialogue as a means to better pursue their professional values. By contrast, the most expensive health care communities featured strong extrinsic incentives. Hospitals and doctors were financially rewarded for each extra output, and doctors viewed meetings with other specialists as a loss of resources. Despite the higher costs, the quality of health was poorer as doctors substituted more services for an integrated treatment plan. A regime change should therefore increase learning, especially double-loop learning through autonomous motivation and allowing professionals to pursue professional values. Autonomous motivation and autonomy have demonstrated a positive relationship with innovation and learning (Hammond et al. 2011), and these factors are expected to improve as a result of the shift towards IL by providing the employees with influence over the goals and increasing their sense of responsibility for attaining the goals. Proposition 2: A shift from an external accountability regime to an internal learning performance regime will decrease gaming behavior and increase cooperation and learning. Finally, a regime change should also affect organizational performance. Our theoretical framework proposes that as EA regimes are relaxed and learning values enhanced, improved behavioral outcomes should result in improved performance outcomes. However, our theory also suggests that EA regimes increase attention to the performance indicators tied to incentives. Overall, then, our theory proposes that changing regimes in the direction of IL accountability should lead to improvements in some aspects of performance but may see declines in performance metrics previously tied to financial rewards. Proposition 3: A shift from an external accountability regime to an internal learning performance regime will increase organizational performance on some aspects but may decrease performance on the previously incentivized performance measures. Conclusion: Making Sense of Contradictions Our approach emphasizes both broad considerations about the design of public services and social science evidence to inform those design choices. By illustrating what an IL regime would entail and by conceiving testable expectations of moving from an EA regime to an IL regime, we have provided an alternative framework to the currently dominant EA regime based on the doctrine of performance management. This approach also allows us to make better sense of some of the contradictions of contemporary governance. One contradiction is the discrepancy between expected and experienced autonomy. The EA regime promises greater organizational autonomy, especially for managers. Yet, in many organizations, opposition to the EA regime is high among both public managers and their frontline employees as they feel constrained and disempowered (Moynihan and Soss 2014). This contradiction makes much more sense when considering that the implementation of the EA regime often takes place in organizations with complex tasks and high levels of autonomous motivation among their professional staff. The EA rationale can conflict with internal professional and service norms and the motivation of employees. Moreover, the posited increase in managerial autonomy is ultimately not delivered. The power to determine the substance of the goals and measures to a large degree determines what happens within the organization. Second, widespread belief in the benefits of the EA regime has driven its introduction in many public services, whereas there is no conclusive evidence of a positive impact on performance. This reflects the broader assessment of Hood and Dixon (2015) that the New Public Management movement of the last 30 years has not made the British central government either cheaper or better. Political principals may favor EA regimes because they imply more control and power over public service production and because of the strong rhetorical appeal of the performance management doctrine. But our model proposes that control may come at the cost of actual improvement in policy outcomes that political principals also care about (Gailmard and Patty 2007). If the expectations developed above about the impact of changing from an EA to an IL regime are valid, the performance effects—except on explicitly measured and incentivized performance measures—would, in complex services, be stronger with an IL regime. The contradictions in current performance regimes call for developing alternative modes of governance. The IL regime provides one such alternative that addresses the challenges with making performance management meaningful in complex public services. On that basis, we propose a clear prescriptive lesson for practice. For complex public services, performance regimes should be centered on learning. In such cases, governments should increase professional autonomy and be wary of high-powered incentives. This does not mean that EA is abandoned but that aspects of EA can threaten professional learning and should therefore be tempered to allow professionals to engage in a goal-based dialogue with political principals. The alternative—to continue to move to systems of EA—can be defended as a political prerogative but not as a model of governing consistent with evidence. The IL regime that we propose may turn out to be no more successful, and need to be modified, but it is time to systematically model and evaluate a different framework for managing public sector performance. References Andersen, Simon C., and Moynihan Donald. 2016. Bureaucratic investments in expertise: Evidence from a randomized controlled field trial. The Journal of Politics 78: 1032–44. Google Scholar CrossRef Search ADS Argyris, Chris. 1980. Making the undiscussable and its undiscussability discussable. Public Administration Review 40: 205–13. Google Scholar CrossRef Search ADS Aucoin, Peter. 1990. Administrative reform in public management: paradigms, principles, paradoxes and pendulums. Governance 3: 115–37. Google Scholar CrossRef Search ADS Baekgaard, Martin, and Serritzlew Søren. 2016. Interpreting performance information: motivated reasoning or unbiased comprehension. Public Administration Review 76: 73–82. Google Scholar CrossRef Search ADS Barzelay, Michael. 2001. The new public management: Improving research and policy dialogue . Berkeley, CA: Univ. of California Press. Barzelay, Michael, and Thompson Fred. 2010. Back to the future: Making public administration a design science. Public Administration Review 70: 295–7. Google Scholar CrossRef Search ADS Behn, Robert D. 2002. The psychological barriers to performance management: Or why isn’t everyone jumping on the performance-management bandwagon? Public Performance & Management Review 26: 5–25. ———. 2014. The PerformanceStat Potential: A leadership strategy for producing results . Washington, DC: Brooking Institution Press. Bevan, Gwyn, and Hood Christopher. 2006. What’s measured is what matters: Targets and gaming in the English Public Health Care System. Public Administration 84: 517–38. Google Scholar CrossRef Search ADS Bevan, Gwyn, and Wilson Deborah. 2013. Does “naming and shaming” work for schools and hospitals? Lessons from natural experiments following devolution in England and Wales. Public Money & Management 33: 245–252. Google Scholar CrossRef Search ADS Bischoff, Ivo, and Blaeschke Frédéric. 2016. Performance budgeting: Incentives and social waste from window-dressing. Journal of Public Administration Research and Theory 26: 344–58. Google Scholar CrossRef Search ADS Blau, Peter M., and Meyer Marshall W.. 1971. Bureaucracy in modern society , 2nd ed. New York, NY: Random House. Bouckaert, Geert, and Halligan John. 2007. Managing performance: International comparisons . London, UK: Routledge. Boyne, George A. 2002. Theme: Local Government: Concepts and indicators of local authority performance: An evaluation of the statutory frameworks in England and Wales. Public Money and Management 22: 17–24. Google Scholar CrossRef Search ADS ———. 2012. Performance management: Does it work? In Public management and performance: Research directions , ed. Richard M. Walker, George A. Boyne, and Gene A. Brewer, 207–26. Cambridge, UK: Cambridge Univ. Press. Google Scholar PubMed PubMed Brehm, John, and Gates Scott. 1999. Working, shirking, and sabotage: Bureaucratic response to a democratic public . Ann Arbor, MI: Univ. of Michigan Press. Brodkin, Evelyn Z. 2011. Policy work: Street-level organizations under New Managerialism. Journal of Public Administration Research and Theory 21:i 253–77. Google Scholar CrossRef Search ADS Campbell, Donald J. 1988. Task complexity: A review and analysis. The Academy of Management Review 13: 40–52. Google Scholar CrossRef Search ADS Campbell, Donald T. 1979. Assessing the impact of planned social change. Evaluation and Program Planning 2: 67–90. Google Scholar CrossRef Search ADS Cots, Francesc, Chiarello Pietro, Salvador Xavier, Castells Xavier, and Quentin Wilm. 2011. DRG-based hospital payment: Intended and unintended consequences. In Diagnosis-related groups in Europe , ed. Reinhard Busse, Alexander Geissler, Wilm Quentin, and Miriam Wiley, 75–92. Berkshire, UK: McGraw-Hill International. Courty, Pascal, and Marschke Gerald. 2003. Dynamics of performance-measurement systems. Oxford Review of Economic Policy 19: 268–84. Google Scholar CrossRef Search ADS ———. 2007. Making government accountable: Lessons from a federal job training program. Public Administration Review 67: 904–16. CrossRef Search ADS Deci, Edward L., Koestner Richard, and Ryan Richard M.. 1999. A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation. Psychological Bulletin 125: 627–68. Google Scholar CrossRef Search ADS Deci, Edward L., and Ryan Richard M.. 2000. The “what” and “why” of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry 11: 227–68. Google Scholar CrossRef Search ADS Dias, Janice J., and Maynard-Moody Steven. 2007. For-profit welfare: Contracts, conflicts, and the performance paradox. J-PART 17: 189–211. Dixit, Avinash. 2002. Incentives and organizations in the public sector: An interpretative review. The Journal of Human Resources 37: 696–727. Google Scholar CrossRef Search ADS Dubnick, Melvin. 2005. Accountability and the promise of performance: In search of the mechanisms. Public Performance & Management Review 28: 376–417. Eterno, John A., and Silverman Eli B.. 2010. The NYPD’s Compstat: Compare statistics or compose statistics? International Journal of Police Science & Management 12: 426–49. Google Scholar CrossRef Search ADS Evetts, Julia. 2003. The sociological analysis of professionalism: Occupational change in the modern world. International Sociology 18: 395–415. Google Scholar CrossRef Search ADS Exworthy, Mark, and Halford Susan. 1999. Professionals and managers in a changing public sector: Conflict, compromise and collaboration. In Professionals and the new managerialism in the public sector , ed. Mark Exworthy and Susan Halford, 1–17. Buckingham, Philidelphia, PA: Open Univ. Press. Figlio, David N., and Getzler Lawrence S.. 2006. Accountability, ability and disability: Gaming the system? Advances in Applied Microeconomics 14: 35–49. Google Scholar CrossRef Search ADS Figlio, David N., and Rouse Cecilia E.. 2006. Do accountability and voucher threats improve low-performing schools? Journal of Public Economics 90: 239–55. Google Scholar CrossRef Search ADS Freidson, Eliot. 2001. Professionalism: The third logic . Chicago, IL: Univ. of Chicago Press. Frey, Bruno S. 2000. Not just for the money: An economic theory of personal motivation . Cheltenham, UK: Edward Elgar. Frey, Bruno S., and Osterloh Margit. 2010. Motivate people with prizes. Nature 465:871. Funnell, Sue C., and Rogers Patricia J.. 2011. Purposeful program theory: effective use of theories of change and logic models . Hoboken, NJ: John Wiley & Sons. Gagné, Maryléne, Forest Jacques, Gilbert Marie-Héléne, Aubé Caroline, Morin Estelle, and Malorni Angela. 2010. The motivation at work scale: Validation evidence in two languages. Educational and Psychological Measurement 70: 628–46. Google Scholar CrossRef Search ADS Gailmard, Sean, and Patty John W.. 2007. Slackers and zealots: Civil service, policy discretion, and bureaucratic expertise. American Journal of Political Science 51: 873–89. Google Scholar CrossRef Search ADS ———. 2013. Learning while governing : Expertise and accountability in the executive branch . Chicago, IL: The University of Chicago Press. Garvin, David A. 2013. How Google sold its engineers on management. Harvard Business Review 91: 74–82. Gawande, Atul. 2009. The cost conundrum. The New Yorker , June 1. Gerrish, Ed. 2016. The impact of performance management on performance in public organizations: A meta-analysis. Public Administration Review 76: 48–66. Google Scholar CrossRef Search ADS Gneezy, Uri, and Rustichini Aldo. 2000. Pay enough or don’t pay at all. The Quarterly Journal of Economics 115: 791–810. Google Scholar CrossRef Search ADS Greve, Henrich R. 2003. Organizational learning from performance feedback: A behavioral perspective on innovation and change . Cambridge, UK: Cambridge Univ. Press. Google Scholar CrossRef Search ADS ———. 2008. A behavioral theory of firm growth. Academy of Management Journal 51: 476–94. CrossRef Search ADS Hammond, Michelle M., Neff Nicole L., Farr James L., Schwall Alexander R., and Xinyuan Zhao. 2011. Predictors of individual-level innovation at work: A meta-analysis. Psychology of Aesthetics, Creativity, and the Arts 5: 90–105. Google Scholar CrossRef Search ADS Hatry, Harry, and Davies Elizabeth. 2011. A guide to data-driven performance reviews . Washington, DC: IBM Center for the Business of Government. Heckman, James J., Heinrich Carolyn, and Smith Jeffrey. 2011. Performance standards and the potential to improve government performance. In The performance of performance standards , ed. James J. Heckman, Carolyn Heinrich, Pascal Courty, Gerald Marschke, and Jeffrey Smith, 1–14. Kalmazoo, MI: WE Upjohn Institute. Google Scholar CrossRef Search ADS Heinrich, Carolyn J. 2007. Evidence-based policy and performance management: Challenges and prospects in two parallel movements. The American Review of Public Administration 37: 255– 77. Google Scholar CrossRef Search ADS Heinrich, Carolyn J., and Marschke Gerald. 2010. Incentives and their dynamics in public sector performance management systems. Journal of Policy Analysis and Management 29: 183–208. Google Scholar CrossRef Search ADS Holmstrom, Bengt, and Milgrom Paul. 1991. Multitask principal-agent analyses: Incentive contracts, asset ownership, and job design. Journal of Law, Economics, & Organization 7 (Special Issue): 24–52. Google Scholar CrossRef Search ADS Hood, Christopher. 2006. Gaming in Targetworld: The targets approach to managing British Public Services. Public Administration Review 66: 515–21. Google Scholar CrossRef Search ADS ———. 2011. The blame game: Spin, bureaucracy, and self-preservation in government . Princeton, NJ: Princeton Univ. Press. Hood, Christopher, and Dixon Ruth. 2015. A government that worked better and cost less? : Evaluating three decades of reform and change in UK central government . Oxford, UK: Oxford University Press. Google Scholar CrossRef Search ADS Hood, Christopher, and Jackson Michael W.. 1991. Administrative argument . Aldershot, UK: Dartmouth. Jacob, Brian A., and Levitt Steven D.. 2003. Rotten apples: An investigation of the prevalence and predictors of teacher cheating. The Quarterly Journal of Economics 118: 843–77. Google Scholar CrossRef Search ADS Jacobsen, Christian B., Hvitved Johan, and Andersen Lotte B.. 2014. Command and motivation: How the perception of external interventions relates to intrinsic motivation and public service motivation. Public Administration 92: 790–806. Google Scholar CrossRef Search ADS Jakobsen, Mads L. F., and Mortensen Peter B.. 2016. Rules and the doctrine of performance management. Public Administration Review 76: 302–12. Google Scholar CrossRef Search ADS Kelman, Steven, and Friedman John N.. 2009. Performance improvement and performance dysfunction: An empirical examination of distortionary impacts of the emergency room wait-time target in the English National Health Service. Journal of Public Administration Research and Theory 19: 917–46. Google Scholar CrossRef Search ADS Kerpershoek, Emiel, Groenleer Martijn, and de Bruijn Hans. 2016. Unintended responses to performance management in Dutch hospital care: Bringing together the managerial and professional perspectives. Public Management Review 18: 417–36. Google Scholar CrossRef Search ADS Kettl, Donald F. 1997. The global revolution in public management: Driving themes, missing links. Journal of Policy Analysis and Management 16: 446–62. Google Scholar CrossRef Search ADS ———. 2015. The transformation of governance: Public administration for the twenty-first century . Baltimore, MD: JHU Press. Koning, Pierre, and Heinrich Carolyn J.. 2013. Cream-skimming, parking and other intended and unintended effects of high-powered, performance-based contracts. Journal of Policy Analysis and Management 32: 461–83. Google Scholar CrossRef Search ADS Kroll, Alexander. 2015. Explaining the use of performance information by public managers: A planned-behavior approach. The American Review of Public Administration 45: 201–15. Google Scholar CrossRef Search ADS Langbein, Laura. 2010. Economics, public service motivation, and pay for performance: Complements or substitutes? International Public Management Journal 13: 9–23. Google Scholar CrossRef Search ADS Latham, Gary P., and Yukl Gary A.. 1975. A review of research on the application of goal setting in organizations. Academy of Management Journal 18: 824–45. Laverty, Stéphane, Lewis David E., and Moynihan Donald P.. 2013. Government reform, political ideology, and administrative burden: The case of performance management in the Bush administration. Public Administration Review 73: 845–57. Google Scholar CrossRef Search ADS Lazear, Edward P., and Michael Gibbs. 2015. Personnel economics in practice (third edition) . Hoboken, NJ: John Wiley & Sons, Inc. Levitt, Barbara, and March James G.. 1988. Organizational learning. Annual Review of Sociology 14: 319–40. Google Scholar CrossRef Search ADS March, James G. 2010. The ambiguities of experience . Ithaca, NY: Cornell Univ. Press. Google Scholar CrossRef Search ADS March, James G., Simon Herbert A., and Guetzkow Harold. 1993. Organizations (second edition) . Cambridge, MA: Blackwell. Mashaw, Jerry. 1983. Bureaucratic justice: Managing social security disability claims . New Haven, CT: Yale Univ. Press. Mathis, William J., and Trujillo Tina. 2016. Lessons from NCLB for the Every Student Succeeds Act . Boulder, CO: National Education Policy Center. http://nepc.colorado.edu/publication/lessons-from-NCLB (accessed May 31, 2017). Mayne, John. 2004. Reporting on outcomes: Setting performance expectations and telling performance stories. The Canadian Journal of Program Evaluation 19: 31–60. Merton, Robert K. 1940. Bureaucratic structure and personality. Social Forces 18: 560–8. Google Scholar CrossRef Search ADS Mintzberg, Henry. 1996. Managing government, governing management. Harvard Business Review 74: 75–83. Moe, Terry. 1989. The politics of bureaucratic structure. In Can the Government Govern? ed. John E. Chubb and Paul E. Petersen, 267–329. Washington, DC: The Brookings Institution. Moynihan, Donald, and Beazley Ivor. 2016. Toward next-generation performance budgeting: Reflections on the experiences of seven reforming countries . Washington, DC: World Bank. Google Scholar CrossRef Search ADS Moynihan, Donald. 2005. Goal-based learning and the future of performance management. Public Administration Review 65: 203–16. Google Scholar CrossRef Search ADS ———. 2008. The dynamics of performance management: Constructing information and reform . Washington, DC: Georgetown Univ. Press. ———. 2010. A workforce of cynics? The effects of contemporary reform on public service motivation. International Public Management Journal 13: 24–34. CrossRef Search ADS Moynihan, Donald, Fernandez Sergio, Kim Soonhee, LeRoux Kelly M., Piotrowski Suzanne J., Wright Bradley E., and Yang Kaifeng. 2011. Performance regimes amidst governance complexity. Journal of Public Administration Research and Theory 21 ( suppl 1): i141–55. Google Scholar CrossRef Search ADS Moynihan, Donald, and Kroll Alexander. 2016. Performance management routines that work? An early assessment of the GPRA Modernization Act. Public Administration Review 76: 314–23. Google Scholar CrossRef Search ADS Moynihan, Donald, and Soss Joe. 2014. Policy feedback and the politics of administration. Public Administration Review 74: 320–32. Google Scholar CrossRef Search ADS Noordegraaf, Mirko. 2007. From “pure” to “hybrid” professionalism: Present-day professionalism in ambiguous public domains. Administration & Society 39: 761–85. Google Scholar CrossRef Search ADS Ordóñez, Lisa D., Schweitzer Maurice E., Galinsky Adam D., and Bazerman Max H.. 2009. Goals gone wild: The systematic side effects of overprescribing goal setting. The Academy of Management Perspectives 23: 6–16. Google Scholar CrossRef Search ADS Ouchi, William G. 1980. Markets, bureaucracies, and clans. Administrative Science Quarterly 25: 129–41. Google Scholar CrossRef Search ADS Pandey, Sanjay K., and Wright Bradley E.. 2006. Connecting the dots in public management: Political environment, organizational goal ambiguity, and the public manager’s role ambiguity. Journal of Public Administration Research and Theory 16: 511–32. Google Scholar CrossRef Search ADS Perry, James L., and Wise Lois R.. 1990. The motivational bases of public service. Public Administration Review 75: 53–78. Perry, James L., Mesch Debra, and Paarlberg Laurie. 2006. Motivating employees in a new governance era: The performance paradigm revisited. Public Administration Review 66: 505–14. Google Scholar CrossRef Search ADS Pollitt, Christopher. 2000. Is the emperor in his underwear? An analysis of the impacts of public management reform. Public Management Review 2: 181–200. ———. 2013. The logics of performance management. Evaluation 19: 346–63. CrossRef Search ADS Pollitt, Christopher, and Bouckaert Geert. 2011. Public management reform: A comparative analysis: new public management, governance, and the neo-weberian state . Oxford, UK: Oxford Univ. Press. Pollitt, Christopher, Harrison Stephen, Dowswell George, Jerak-Zuiderent Sonja, and Roland Bal. 2010. Performance regimes in health care: Institutions, critical junctures and the logic of escalation in England and the Netherlands. Evaluation 16: 13–29. Google Scholar CrossRef Search ADS Roberts, Jennifer, and Dietrich Michael. 1999. Conceptualizing professionalism. American Journal of Economics and Sociology 58: 977–98. Google Scholar CrossRef Search ADS Romzek, Barbara S., and Dubnick Melvin J.. 1987. Accountability in the public sector: Lessons from the Challenger Tragedy. Public Administration Review 47: 227–38. Google Scholar CrossRef Search ADS Ryan, Richard M., and Weinstein Netta. 2009. Undermining quality teaching and learning: A self-determination theory perspective on high-stakes testing. Theory and Research in Education 7: 224–33. Sauder, Michael, and Espeland Wendy N.. 2009. The discipline of rankings: Tight coupling and organizational change. American Sociological Review 74: 63–82. Google Scholar CrossRef Search ADS Scharpf, Fritz W. 1994. Games real actors could play: Positive and negative coordination in embedded negotiations. Journal of Theoretical Politics 6: 27–53. Google Scholar CrossRef Search ADS Simon, Herbert A. 1947. A comment on “The Science of Public Administration.” Public Administration Review 7: 200–3. Google Scholar CrossRef Search ADS Soss, Joe, Fording Richard C., and Schram Sanford. 2011. Disciplining the poor: Neoliberal paternalism and the persistent power of race . Chicago, IL: Univ. of Chicago Press. Google Scholar CrossRef Search ADS Steiber, Annika. 2014. The Google model: Managing continuous innovation in a rapidly changing world . Cham, Switzerland: Springer International Publishing. Stout, Margaret. 2010. Revisiting the (lost) art of ideal-typing in public administration. Administrative Theory & Praxis 32: 491–519. Google Scholar CrossRef Search ADS Suddaby, Roy, Cooper David J., and Greenwood Royston. 2007. Transnational regulation of professional services: Governance dynamics of field level organizational change. Accounting, Organizations and Society 32: 333–62. Google Scholar CrossRef Search ADS Søgaard, Rikke, Kristensen Søren R., and Bech Mickael. 2015. Incentivising effort in governance of public hospitals: Development of a delegation-based alternative to activity-based remuneration. Health Policy 119: 1076–85. Google Scholar CrossRef Search ADS Sørensen, Marie-Louise L., and Burau Viola. 2016. Why we need to move beyond diagnosis-related groups and how we might do so. Journal of Health Services Research & Policy 21: 64–6. Google Scholar CrossRef Search ADS Talbot, Colin. 2010. Theories of performance: Organizational and service improvement in the public domain . Oxford, UK: Oxford Univ. Press. The New Yorker. 2016. How regulation failed with Wells Fargo. The New Yorker , September 12. Tummers, Lars G., and Van de Walle Steven. 2012. Explaining health care professionals’ resistance to implement diagnosis related groups: (No) benefits for society, patients and professionals. Health Policy 108: 158–66. Google Scholar CrossRef Search ADS van Loon, Nina, Leisink Peter M., Knies Eva, and Brewer Gene A.. 2016. Red Tape : Developing and validating a new job-centered measure. Public Administration Review 76: 662–73. Google Scholar CrossRef Search ADS Weber, Max. 1960. The essentials of bureaucratic organization: An ideal-type construct. In Reader in bureaucracy , ed. Robert K. Merton, 2nd printing, 18–27. Glencoe, IL: The Free Press. Weibel, Antoinette, Rost Katja, and Osterloh Margit. 2010. Pay for performance in the public sector: Benefits and (hidden) costs. Journal of Public Administration Research and Theory 20: 387–412. Google Scholar CrossRef Search ADS Wright, Bradley E., and Pandey Sanjay K.. 2010. Transformational leadership in the public sector: Does structure matter? Journal of Public Administration Research and Theory 20: 75–90. Google Scholar CrossRef Search ADS Footnotes 1 A variety of administrative processes could be categorized to fall into this conceptualization. For example, Mashaw (1983) portrays administrative law judges juggling adjudication metrics with some professionally-based notion of bureaucratic justice and devising learning mechanisms to bridge the two. 2 In contrast to both Talbot’s (2010) and Pollitt’s (2013) broader performance regime and management perspectives with multiple actors and interventions, we make the degree of externally imposed goals and incentives the key dimension on which we distinguish between different performance regimes. Another difference to prior work that has critiqued what we label the EA regime (e.g., Aucoin 1990), is that we explicitly apply a learning perspective, though note that Mashaw (1983) focused on elements of learning in his conceptualization of “internal administrative law” for disability claims. © The Author 2017. Published by Oxford University Press on behalf of the Public Management Research Association. All rights reserved. For permissions, please e-mail: email@example.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)
Perspectives on Public Management and Governance – Oxford University Press
Published: Sep 27, 2017
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera