TY - JOUR
AU - Feiock, Richard, C
AB - Abstract A normative assumption of government reform efforts such as New Public Management is that fostering a more innovative, proactive, and risk-taking organizational culture—developing what has been described as an “entrepreneurial orientation” (EO)—improves performance. But in arenas like urban sustainability, performance can be an ambiguous, multifaceted concept. Managers’ assessments of their own nimbleness, innovative thinking, and risk culture are also likely to influence how they interpret the risk-reward balance of opportunities to enhance organizational performance. This study examines how meso-level organizational decisions impact managers’ individual risk-assessments of sustainability initiatives. We do so through a combination of Bayesian structural equation modeling of US local government survey data collected over two time periods, and an artifactual survey experiment with empaneled local government employees. This multimethod design allows us to examine the role of organizational performance and EO—meso-level learning heuristics—in shaping the micro-foundations of managerial risk assessment. The organization-level observational results indicate that local governments engage in risk-seeking behavior in order to minimize their potential for losses of prior effort. Experimental results confirm local government administrators are loss-averse when asked to evaluate the merits of initiating a hypothetical sustainability program. Introduction When “bureaucrat” enters the mind of the average citizen, the image likely summons a litany of popular pejoratives: lazy, incompetent, robotic, or risk averse (Brinton Milward and Rainey 1983; Lee and Van Ryzin 2020). These stereotypes overlook the “wickedness” of problems, complexity of governmental programs, and decision-making constraints public managers operate within (Head and Alford 2015; Kettl 2016; Rutherford and Meier 2015). For instance, studies of intergovernmental programs demonstrate how disconnects between policy design and implementation can delay or thwart achievement of program objectives (Pressman and Wildavsky 1973; Terman et al. 2016). Policy designers can have incongruent goals (Nicholson-Crotty 2004), or fail to provide adequate technical guidance to managers (Carley, Nicholson-Crotty, and Fisher 2015). In such scenarios, managers must work with incomplete information (March and Simon 1958). However, like all individuals, they also suffer from cognitive biases which skew decision making and amplify through organizational choices (Meier, Favero, and Zhu 2015; Nicholson-Crotty, Nicholson-Crotty, and Fernandez 2017). This study explores a latent cognitive driver of bias in public managerial decision making—aversion to loss (Battaglio et al. 2018). Organizations are thought to weigh losses differently when evaluating the risks of continuing existing programs versus initiating new ones (March and Simon 1958). Decades of research in psychology and behavioral economics has extended bounded rationality to consumer, employment, and health choices (Kahneman 2011; Kahneman and Tversky 1986; Redelmeier, Katz, and Kahneman 2003; Thaler and Ganser 2015). In particular, prospect theory pioneered by Kahneman and Tversky (1979) holds that variation in how potential losses and gains are semantically “framed,” systematically alters preferences (Tversky and Kahneman 1981). Potential losses outweigh potential gains when choices are presented under risk and uncertainty (Kahneman and Tversky 2013). A normative assumption of government reform efforts such as New Public Management (NPM) is that fostering a more innovative, proactive, and risk-taking organizational culture—or developing what has been described as an “entrepreneurial orientation” (EO)—can enable managers to be more innovative problem-solvers (Bellone and Goerl 1992; Osborne and Gaebler 1992; Swann 2017). We know less about whether managers’ nimbleness, innovative thinking, and risk-taking are influenced by loss aversion when assessing risky opportunities. We examine loss aversion in the context of one of the more complex policy quagmires local governments confront today—sustainability. Urban sustainability is often defined as efforts by local governments to balance of environmental, equity, and economic aims within communities (Mazmanian and Kraft 2009; Portney 2013; Slavin 2013), and has gained prominence as national governments have reneged on pledges to confront climate change (Deslatte 2020a; Hughes 2019). Local government managers venturing into sustainability confront inherently ambiguous, high-risk alternatives, given the often conflicting economic, ecological, and social values or goals of citizens and elected officials (Zeemering 2014). Thus, our central research questions pertain to establishing the micro-foundations of managerial risk assessment in a complex, ill-defined context. How are organizational decisions about sustainability programs shaped by learning heuristics, and how does organizational learning impact managers’ individual risk-assessments? We address these questions through a multimethod research design that examines the role of organizational performance and EO at both the organizational (meso) and individual (micro) levels (Hendren, Luo, and Pandey 2018; Mele and Belardinelli 2019). First, we conduct an observational analysis drawing from two national surveys of US local government experiences in a federally initiated energy-efficiency program. We then explicitly test for the causal mechanism of loss aversion using an artifactual survey experiment which manipulates the framing of a potential new methane capture and reuse program. Synthesizing across both analyses, we find evidence that local governments are more likely to take risks to avoid losses than to achieve gains. When deciding whether to continue funding their local energy-efficiency efforts, local government managers’ dissatisfaction with the program was overwhelmed by positive perceptions of the organization’s overall sustainability performance. This positive perception of the organization appeared to drive decisions to continue financially supporting the program, consistent with the “sunk costs of innovation” argument offered by March and Simon (1958, 194). Thus, local governments engaged in risk-seeking behavior in order to minimize the potential for losses of prior effort. However, under an alternative experimental condition, local government administrators were risk averse when evaluating the merits of creating a new, hypothetical sustainability program. Behavioral Underpinnings of an Organizational Theory of Choice In organizational science, meso- and micro-approaches offer complimentary—although, not always combinatory—approaches for assessing the normative belief that governmental reforms can address the existential threats facing society (Anechiarico 2007; Fiorino 2018; Ingraham 2005; Kettl 2016). They both also rest on a foundational assumption that information processing, or learning, generally does not occur in a fully rational way, but rather through organizational learning heuristics, trial and error, motivated reasoning, and other cognitive short cuts. In a similar vein, the literature on policy learning emphasizes how organizations and individuals utilize a “learning capacity” to adjust and steer policy change over time (Deutsch 1965; Dunlop and Radaelli 2013; Moyson, Scholten, and Weible 2017). Collectively, these literatures pay varying degrees of attention to the poorly understood linkages between individual and organizational learning processes (Dunlop and Radaelli 2017; Fernandes and Simon 1999; Jones 2001). This article leverages meso- and micro-approaches to zero in on managerial risk-assessments of organizational initiatives (Battaglio et al. 2018; Bullock, Greer, and O’ Toole 2019). To do so, we explicitly examine organizational learning heuristics and how they shape individual choices in the context of sustainability. We first examine what led organizations to make sometimes sizable commitments to an ongoing, sustainability effort instigated by a massive federal intervention, the US Department of Energy’s (DOE) $3.2 billion Energy Efficiency and Conservation Block Grant Program (EECBG). Then, we examine the linkages between meso-level perceived organizational performance, EO, and the micro-level latent cognitive mechanism of loss aversion under experimental conditions. Through a historical choice cities made and a hypothetical one managers were asked to make, this approach provides a more comprehensive examination into the cognitive limitations of managers tasked with assessing risk and making programmatic choices. Organizational Performance and the Sunk Cost of Innovation Local government management entails risk management. In recent decades, a mantra has emerged that cities should meet challenges by thinking and acting more entrepreneurially (Audretsch 2015; Florida, Adler, and Mellander 2017; Teske and Schneider 1994). Management scholars have operationalized entrepreneurial orientation (EO) through dimensions of organization-level proactiveness, innovativeness, and risk taking (Meynhardt and Diefenbach 2012). More recently, behavioral public administration researchers have sparked wider interest in micro-level approaches focused on the cognitive biases of public managers, citizens and politicians, such as anchoring (Feeney 2012), confirmation (Grimmelikhuijsen and Meijer 2012), status quo (Jilke, Van Ryzin, and Van de Walle 2016), and framing effects (Belardinelli et al. 2018). An implicit assumption in EO research is that managers can use public resources to capitalize on social opportunities (Morris and Jones 1999). An implication of behavioral public administration is that managers may not be so good at sorting out promising opportunities from risky gambits (Battaglio et al. 2018; Nicholson-Crotty, Nicholson-Crotty, and Webeck 2019). This is an important problem because risk assessment is central to government reform movements. Assessing risk is at the heart of performance management and strategic management reforms (George et al. 2020). The literature on public-sector entrepreneurialism, in particular, stakes a position that risk-taking is necessary for being more entrepreneurial and improving performance (Kearney and Meynhardt 2016; Morris and Jones 1999). However, performance management and strategic management are premised on helping organizations better assess risks and rewards through a bundle of learning heuristics which can be inputs to any organizational decision process (Heclo 2010; Moynihan 2009). Building from behavioral foundations, Bullock, Greer, and O’ Toole (2019) outline a research agenda for advancing a theoretically informed view of risk in public management. In particular, they call for greater attention to the role of loss aversion. Based in social psychology, prospect theory has demonstrated that loss aversion results when perceived gains from choices under risk and uncertainty are outweighed by potential losses (Tversky and Kahneman 1981). Applied to organizations, prospect theory holds that the expected utility of an initiative under consideration is relative to some reference point. Since boundedly rational organizations rely on learning heuristics (Jones 2002; Simon 1991), one such heuristic is the gap between past performance and expectations (Meier, Favero, and Zhu 2015; Rutherford and Meier 2015). One recent model of relative risk aversion suggests public managers’ risk preferences are contingent on past performance, varying depending on where the organization falls relative to its aspirational standards (Nicholson-Crotty, Nicholson-Crotty, and Fernandez 2017). A behavioral approach to organizational risk-taking suggests organizations have difficulty making changes, and condition their programmatic decisions less on a comprehensive evaluation of all the potential options than on their recent experiences, agenda-setting efforts, and emotions (Jones 2003). March and Simon called this failure to seek out alternatives the “sunk cost of innovation” (1958, 194), arguing choices between program continuation and initiation are not evaluated equally. Rather, organizations give preference to programs already in existence compared to initiating change. All else being equal, program continuity tends to prevail, while search is reduced as satisfaction increases (March and Simon 1958). Loss aversion manifests in this context as a greater willingness by organizations to expend resources to avoid the losses of discontinuing a program than to achieve gains through innovation (Kahneman and Tversky 2013). Perceptions of performance and EO may both offer contrasting heuristics for making such decisions. Past performance provides one cognitive short-cut for evaluating future resource commitments. EO, as codified through decision-processes, process realignment, and beliefs, offers another. More entrepreneurial organizations potentially make stronger semantic associations between risk and performance, and thus have more stable preferences for taking risks. A meso–micro approach for examining this phenomenon first requires alternative scenarios of program continuation and initiation. Local governments have accumulated decades of experience and programmatic commitments to enhancing sustainability via energy, climate, land use, waste, natural resource, social, and infrastructure goals and objectives (Deslatte and Swann 2016; Swann and Deslatte 2019). Many of these efforts meet the criteria for complex, ill-defined problems (Funke 2001), in that not all variables and outcomes are directly observable; multiple goals may conflict with each other; patterns of connectivity between variables are present; and many outcomes, such as mitigation of climate change, are time delayed (Fernandes and Simon 1999). Like many goals with uncertain outcomes, long-term sustainability problems such as climate risk often take a back seat to more amenable, quantifiable, and solvable ones such as reducing energy costs. In short, sustainability represents a policy arena characterized by “sunk costs of innovation.” As such, we expect organizations tend to prioritize problems in this arena and make choices, which favor continuity and certainty. Managerial Loss Aversion and Equivalency Framing Program initiation is an inherently high-risk proposition. Public managers are likely to respond to choice scenarios differently depending on how the risk of the situation is framed. Prospect theory has established that individuals tend to make less-risky decisions when their choices are framed in terms of options between different “gains” scenarios, and they will take greater risks when choices are framed between probable and determinant losses (Kahneman and Tversky 1986). In other words, individuals will take greater risks to avoid losses than they will to achieve greater gains (Kahneman 2011). This inconsistency in choices under uncertainty has fueled a voluminous economics and political science literature on framing effects (Druckman 2001). In public administration, these studies have tended to focus on equivalency framing, in which different but logically equivalent—often numeric—descriptions of performance are presented (James, Jilke, and Van Ryzin 2017; Olsen 2015; Piotrowski, Grimmelikhuijsen, and Deat 2019). For instance, Grosso, Charbonneau, and Van Ryzin (2017) found that citizens were more supportive of an HIV prevention program when shown outputs, the number of people served, as opposed to outcomes, or infections prevented (Grosso, Charbonneau, and Van Ryzin 2017). Loss aversion is the tendency to overweight the potential for losses relative to equivalent gains. Public managers—just like any other individuals—may demonstrate greater risk tolerance when their choices are famed as potential losses. Recently, Nicholson-Crotty, Nicholson-Crotty, and Webeck (2019) found by replicating a series of widely used framing experiments that public- and private-sector managers had higher risk tolerance when their alternatives for responding to a hypothetical disease outbreak were framed in terms of population losses as opposed to lives saved (Nicholson-Crotty, Nicholson-Crotty, and Webeck 2019). A behavioral theory of organizational choice requires linking performance and risk assessment through elements common to a model of both organizational and individual decisions (Bullock, Greer, and O’ Toole 2019; Jones 2003). For instance, Simon’s model of bounded rationality includes commonalities between how organizational and individual memories are coded into rules, the limits of attention present in agenda-setting, disproportionate information processing, and the role of emotion in setting priorities (Cohen, March, and Olsen 1972; Jones and Mortensen 2018; Kingdon and Thurber 1984; March and Simon 1958; Simon 1996). Individual managers should base their own decisions on prior organizational learning which has been codified, although neither individuals nor organizations can sort all available information and must rely on heuristics (Deslatte 2018; Jones 2001). This information-processing occurs cyclically, through what the policy learning literature calls “single-loops” and “double-loops” in which organizations adopt their objectives and then modify them as errors are detected and corrected (Moyson, Scholten, and Weible 2017). However, organizations do not always have the option of time and “double-loop” learning when committing resources. Managers can also identify emotionally with organizational goals, allowing ideologies, professional norms, or prepackaged “best practices” to guide solutions for high-risk problems (Fernandes and Simon 1999). Bounded rationality implies that managerial risk assessment should suffer from limits of attention and cognitive capabilities, but be causally connected to organizational learning processes. Taking a meso–micro approach to program initiation, public managers take their cues from past organizational experiences and learning either encoded in hardware (rules, regulations, procedures) or software (culture). But they should display systematic risk aversion when the potential of a new program is framed in terms of its gains. A Choice Context: Local Sustainability Program Performance The analysis first examines the meso-level linkages between organizational learning and program continuation through the context of a US federal program which instigated thousands of local sustainability initiatives. The EECBG program, implemented as part of the American Recovery and Reinvestment Act (ARRA) of 2009, financed roughly 7,400 projects to increase energy efficiency and reduce fossil fuel emissions in local communities (Terman et al. 2016). Under the EECBG program, local governments could conduct energy audits of government and residential buildings; create energy-efficiency programs and incentives; retrofit buildings; construct bike lanes; start recycling programs; replace street-lighting; install technologies to reduce methane or greenhouse gas emissions from government operations, among many other activities (DNV GL 2015). A large literature on intergovernmental implementation has posited that clear objectives and guidance, agreement on goals, and the capacities of the implementing agencies work in tandem to shape outcomes (Goggin 1990; Mazmanian and Sabatier 1983). Local governments tend to lack the technical and fiscal capabilities to navigate complicated federal grant application and administration processes, which is why upfront technical assistance from principals—particularly in one-shot grants such as the EECBG program—is critical to minimizing errors (Terman and Feiock 2015; Terman et al. 2016). In this case, both state and local governments reported unclear guidance from federal officials and the ARRA legislation and EECBG program both suffered from implementation delays (Carley, Nicholson-Crotty, and Fisher 2015; Jennings, Hall, and Zhang 2012; Terman et al. 2016). One of the criteria DOE used for evaluating applicants was the likelihood they would continue their activities after the federal grant had expired (Deslatte 2019). We argue this decision to replace federal one-time funding with recurring, local revenues required risk assessment. Prospect theory suggests that the risk of losing programmatic benefits outweighs the potential for gaining equivalent benefits from program expansion. A decision to nonrenew activities or initiatives started through EECBG grants creates the potential for losses of benefits, and it is reasonable to assume based on the implementation literature that past experiences likely play some role in such evaluations (Carley, Nicholson-Crotty, and Fisher 2015). Several aspects of bounded rationality are applicable to this decision. Agenda-setting involves assessing stimuli from the environment and classifying them as relevant or irrelevant for decision making, a process which has been compared to an individual’s short-term memory (Jones 2002, 2003). Local governments implementing EECBG-funded activities likely encountered more program-relevant information than what they deemed necessary to make implementation and evaluation decisions. The effect of agenda-setting here is that some performance-relevant data are more heavily weighted than others. Direct experiences with the program are one source of information which may have influenced an organization’s choice to continue the program. Since many of the activities eligible for EECBG funding were new to city governments, they required staff and training. Cities made extensive use of contractors, but the contracted services were subject to federal labor and contracting requirements, thus investments in contract management capacity were necessary (Terman et al. 2016). The EECBG program also encountered implementation delays, technical problems with online reporting and monitoring, as well as local governments’ lack of staffing and expertise navigating federal grants (Terman and Feiock 2015; GAO 2011). These experiences may have drawn additional attention—serial processing, or the act of attending to one item or issue at a time—as existing grant-management routines proved inadequate or administrative resources were found lacking. “We were able to do a lot of great projects with the money, but the reporting and close out were horribly time consuming,” one local government department manager wrote in response to a 2013 survey of EECBG recipients. While negative experiences with the program may have prompted some cities to abandon their EECBG initiatives, others reported organizational learning and increased buy-in from the experience. “Installing LED streetlights with EECBG funds gave our city confidence to pursue a much larger installation of LED streetlights on our own,” one manager wrote. “EECBG broke the ice and has been critical in that effort.” The result is that while grant management is a core function of local governments, direct experience with the EECBG program likely stands out as a learning experience encoding new solutions into organizational memory. These new organizational memories could increase commitment for continuing the program after federal funds expired. Thus, we expect that serial processing—triggered by problems encountered with the program—fostered a greater intent to codify efforts, and subsequently, an aversion to wasting these efforts and resources. The EO literature suggests that managers’ assessments of their organization’s own nimbleness, innovative thinking, and risk culture are also likely to influence this choice. More entrepreneurial managers are more likely to believe that greater risks are necessary to enhance performance and have been shown to proactively engage stakeholders, and develop technical and fiscal capacities to further performance innovations (Deslatte and Swann 2020). If this is the case, then an organizational EO should foster a willingness to take risks irrespective of other contradictory or unclear information signals. We expect that direct experience with the program—satisfaction—can interact with perceptions of EO. In this instance, organizations weighed their program experiences against their sense of EO when they considered whether to extend the EECBG program. This suggests that managers who work in organizations they considered more risk-taking, innovative, or proactive may be willing to overlook negative experiences with EECBG implementation or administration. Organizations with higher levels of EO should be less affected by low satisfaction, evidenced by a positive interaction effect. We utilize an interactive-term hypothesis (Berry, Golder, and Milton 2012) to state this expectation more plainly: H1: Local government EECBG satisfaction will have a positive, direct effect on commitment to EECBG program continuation at all levels of EO; at lower levels of satisfaction, the moderating effect of EO strengthens. Another form of information impacting choice is an organization’s perceived overall or diffuse assessment of their sustainability performance. This is premised on the information-processing theory of policy dynamics advanced by Jones, Baumgarter, Workman, and others (Jones and Baumgartner 2012; Workman, Jones, and Jochim 2009). Most of the recipients of EECBG funds either had some previous experience with economic, environmental, or social sustainability efforts, or were initiating sustainability planning and activities beyond those funded through the program (Terman and Feiock 2015). In other words, they had routinized the processing of this information throughout organizations. Unlike the serial processing which may occur with experiences from a single program, diffuse assessments lean on parallel processing, given that activities or experiences are spread throughout an organization. But organizations suffer from similar limits as individuals in allocating attention when information is oversupplied (Jones 2001). In such assessments, organizational leaders may rely on heuristics, reputation, or assumptions about organizational performance. These diffuse experiences are also more likely to influence emotional contagion, which is the arousal of employees’ attention to the appropriateness or importance of an organization’s mission (Jones 1994). Emotional contagion can be useful for motivating employees, but can override a rational benefit-cost analytic decision process and make it more difficult to change course (Krause, Yi, and Feiock 2016). Employees who feel strongly about their organization’s overall track record and commitment to sustainability goals may be willing to look past poor individual programmatic experiences. As a consequence, diffuse experiences with organization-wide sustainability efforts and perceptions of performance may bypass or override serially-processed information specific to the EECBG program. The net effect is that organizations with more positive, diffuse experiences with sustainability should be more averse to risks of loss and more likely to continue the EECBG program irrespective of their direct experience with it. As previously noted, loss aversion suggests governments should be more willing to preserve existing programs in order to avoid the losses incurred than expand into new ones in the hope of future gains. However, the government reform mantra suggests that organizations with a higher EO will have a more stable tolerance for risk irrespective of past performance. We test these somewhat rivalrous expectations by conditioning the relationship between performance and risk-taking in the form of program continuation on the level of EO. In other words, we expect that EO moderates the risk calculus of organizations. Performance should have a positive effect on risk-taking but weaken at high levels of EO, and vice versa. We again utilize an interactive-term hypothesis to state this expectation: H2: Perceptions of organization-wide sustainability performance will have a positive, direct effect on commitment to EECBG program continuation at all levels of EO; at high levels of performance, the moderating effect of EO weakens. The linkages between organizational learning and organization decision-making assume some causal connection exists between meso- and micro-level information processing. However, mapping organizational information-processing heuristics to micro-foundations is an elusive empirical phenomenon in both the public management and policy process literatures (Bullock, Greer, and O’ Toole 2019; Dunlop and Radaelli 2017; Moyson, Scholten, and Weible 2017). Thus, the second stage of our analysis examines the influence of meso-level learning on micro-level decisions in an experimental program initiation scenario. While our prior organization-level EECBG program context reflects a historical risk-taking effort, it cannot disentangle managerial loss aversion as a causal mechanism from other elements of organizational learning which may also influence the choice. That said, public managers clearly play a role in policy choices (Carr 2015; Deslatte, Swann, and Feiock 2017). Following recent work on public managers’ loss aversion (Nicholson-Crotty, Nicholson-Crotty, and Webeck 2019), we examine support for sustainability program scenarios in which public-sector workers must consider the pros and cons of allocating public resources to achieve public purposes. Thus, we hypothesize: H3: Local government employees will be more risk-seeking when new sustainability program costs and benefits are framed in terms of losses as opposed to equivalent gains. The final proposition we explore returns again to whether an organizational EO influences the perceptual weights managers assign to opportunities for gains and losses—and thus, their risk behavior—relative to sustainability initiatives. The government reform agenda of recent decades assumes managers can be more risk-taking, although normative theories such as NPM do not specify the types of risk or conditions under which risk should be minimized (Bullock, Greer, and O’ Toole 2019). There is some recent evidence which suggests public administrators are more willing to be entrepreneurial when facing immediate resource constraints (Singla, Stritch, and Feeney 2018). However, no studies of which we are aware examine whether EO mitigates or exacerbates loss aversion. Identification with an organization’s mission is one element of bounded rationality relevant for this question. Because the literature does not delve into the appropriateness of risk-taking, we test whether individuals in organizations with higher EO are more risk-taking regardless of whether opportunities are framed as equivalent losses or gains. H4: EO will be positively associated with risk-seeking when new sustainability program costs and benefits are framed in terms of both gains and losses. Research Design Our study draws data from multiple sources in order to test whether the cognitive mechanism of loss aversion is more likely to influence the willingness of local governments to engage in sustainability initiatives. We utilize two inferential approaches that focus observationally on the organizational level and experimentally on the individual levels. Organizational (Meso) Analytic Method The organizational-level analysis uses responses from two national surveys of US local governments which received EECBG funding for sustainability programs. The EECBG program, which distributed $3.2 billion in funds from the American Recovery and Reinvestment Act (ARRA) of 2009, financed more than 7,400 projects undertaken by 2,187 state, local, and tribal governments. The surveys were administered in 2012 and 2013 by the Local Governance Research Laboratory at Florida State University and are available from the authors upon request. The 2012 survey asked local government managers designated as the official liaison to the federal government a battery of questions assessing the EO of organizations, their satisfaction with the program, and their overall organizational performance on sustainability goals. The 2012 survey was sent to all cities and towns that received EECBG funds. A response rate of 67% was achieved. Respondents covered 47 states and Puerto Rico, with a median population of 67,287, which tilts in the direction of larger municipalities due to the stipulation that grant recipients needed a population greater than 35,000, although smaller municipalities could still apply and obtain funds through their state government. The 2013 follow-up survey was sent to the same sampling frame of cities, asking local government managers whether and how they planned to continue their programs or initiatives after their federal grant funds were exhausted. It generated a response rate of roughly 53%, or 513 out of 968 eligible city recipients. Merging responses of local governments which responded to both surveys yielded an N = 338. The median population for the combined survey was 69,272, and the 2012 and merged 2012–13 samples were not substantively different in terms of racial composition, median household income, or educational achievement. To mitigate potential for common source bias (Favero and Bullock 2015), we draw our independent and dependent variables from the separate 2012 and 2013 surveys, respectively. The dependent variable, commitment to EECBG program continuation (commit), was measured with responses to the 2013 survey (T2), which asked: “For any of the program categories below that were supported by EECBG grant funds, please indicate what happened or what you anticipate will happen with the termination of the EECBG grant.” We used 10 survey items, covering a range of energy efficiency, green infrastructure, and conservation programs (Supplementary Appendix A). To capture commitment to program continuation, we collapsed two survey responses, “service not funded with EECBG” and “most will be terminated,” and set them equal to 0. We then set the responses “most will be continued at reduced level” equal to 1, and “most will be continued at same level” equal to 2. The construct demonstrated high reliability (α = .89). Three independent variables were developed from responses to the 2012 survey (T1). We measured EO with five 5-point Likert-type scale items (1 = strongly disagree, 3 = neutral, 5 = strongly agree) related to governments’ perceived risk-taking; innovativeness; bureaucratization (reversed); growth and resource acquisition; and dynamism and entrepreneurialism (α = .88). We captured perceived performance (perform) with responses to the following question: “How would you describe your city’s performance in the past two years on the following?” Five items measured on a 4-point Likert-type scale (1 = situation is worse, 2 = no change, 3 = some positive effects, 4 = big positive effects) were used, including reduction in government energy use; cleaner environment; green job creation; greenhouse gas reduction; and financial savings from sustainability (α = .79). We measured EECBG program satisfaction (satisfied) with five 10-point Likert-type scale items, including satisfaction with funding disbursement process; technical assistance; DOE project oversight; performance/status reporting; and overall satisfaction with EECBG (α = .90). Informed by the organizational task environment literature (Dess and Beard 1984; Walker, Berry, and Avellaneda 2015), we controlled for local governments’ capacity for continuing sustainability programs with a perceptual measure developed from the 2012 survey: an eight-item construct for organizational capacity to implement sustainability programs (capacity) (α = .79). Capacities are important predictors of city sustainability activity (Wang et al. 2012). Our measure captures the degree to which financial, managerial, and technical capacities such as costs/budget conflicts, contracting and procurement delays, staff capacity, availability of qualified contractors, and informational resources, as well as support from local elected officials and community-based groups, facilitated or hindered the implementation of EECBG programs. Additional capacity indicators included logged population (ln_pop) and logged median family income (ln_inc) from the US Census Bureau, as well as per capita own source revenue in the $100s (pc_osr) from the US Census of Governments, 2007. We also included a dichotomous measure indicating whether a city belonged to the network ICLEI-Local Governments for Sustainability (ICLEI) or otherwise. To capture complexity, we calculated Herfindahl–Hirschman Indices (HHIs) for city-level age (age_hhi) and racial/ethnic (eth_hhi) diversity, with higher scores indicating greater diversity and thus more complexity. For dynamism, we controlled for the population change from 2000 to 2010 (pop_ch). Generally, we expect local governments with greater capacity and dynamism, and less complexity, to be more likely to commit to continuing sustainability programs after EECBG funds have been expended. Table 1 reports the descriptive statistics and bivariate correlations for all variables in the model. Table 1. Descriptive Statistics and Zero-Order Correlations . Variable . Mean . SD . 1 . 2 . 3 . 4 . 5 . 6 . 7 . 8 . 9 . 10 . 11 . 1 Commit 0.66 0.8 2 EO 3.19 1.01 0.13* 3 Perform 2.97 0.6 0.32* 0.18* 4 Satisfied 6.95 2.49 0.15* −0.02 0.14* 5 Capacity 4.14 1.09 −0.07 0.22* −0.02 0.15* 6 Ln_pop 11.27 0.86 0.17* −0.11* 0.31* 0.18* −0.11 7 Ln_inc 11.04 0.30 −0.11 0.12* 0.06 −0.14* 0.03 −0.1 8 Pc_osr 5.74 3.88 0.02 0.01 0.07 0.02 0.05 0.06 0.1 9 ICLEI 0.29 0.45 0.17* −0.02 0.29* 0.03 −0.07 0.32* 0.01 0.07 10 Age_hhi 0.37 0.03 0.05 0.14* 0.03 0.08 0.09 −0.11* 0.31* 0.05 −0.05 11 Eth_hhi 0.53 0.15 −0.13* 0.09 −0.17* −0.1 0.12* −0.3* 0.12* −0.06 −0.12* −0.12* 12 Pop_ch 27.88 229.64 −0.05 0.06 −0.12* −0.02 0.09 −0.04 0.03 −0.01 −0.04 0.09 −0.05 . Variable . Mean . SD . 1 . 2 . 3 . 4 . 5 . 6 . 7 . 8 . 9 . 10 . 11 . 1 Commit 0.66 0.8 2 EO 3.19 1.01 0.13* 3 Perform 2.97 0.6 0.32* 0.18* 4 Satisfied 6.95 2.49 0.15* −0.02 0.14* 5 Capacity 4.14 1.09 −0.07 0.22* −0.02 0.15* 6 Ln_pop 11.27 0.86 0.17* −0.11* 0.31* 0.18* −0.11 7 Ln_inc 11.04 0.30 −0.11 0.12* 0.06 −0.14* 0.03 −0.1 8 Pc_osr 5.74 3.88 0.02 0.01 0.07 0.02 0.05 0.06 0.1 9 ICLEI 0.29 0.45 0.17* −0.02 0.29* 0.03 −0.07 0.32* 0.01 0.07 10 Age_hhi 0.37 0.03 0.05 0.14* 0.03 0.08 0.09 −0.11* 0.31* 0.05 −0.05 11 Eth_hhi 0.53 0.15 −0.13* 0.09 −0.17* −0.1 0.12* −0.3* 0.12* −0.06 −0.12* −0.12* 12 Pop_ch 27.88 229.64 −0.05 0.06 −0.12* −0.02 0.09 −0.04 0.03 −0.01 −0.04 0.09 −0.05 Means and SD were calculated by creating averaged scores across the items that comprised the constructs, although such scales were not used in the SEM analysis and are displayed for illustration purposes only. *p < .05. Open in new tab Table 1. Descriptive Statistics and Zero-Order Correlations . Variable . Mean . SD . 1 . 2 . 3 . 4 . 5 . 6 . 7 . 8 . 9 . 10 . 11 . 1 Commit 0.66 0.8 2 EO 3.19 1.01 0.13* 3 Perform 2.97 0.6 0.32* 0.18* 4 Satisfied 6.95 2.49 0.15* −0.02 0.14* 5 Capacity 4.14 1.09 −0.07 0.22* −0.02 0.15* 6 Ln_pop 11.27 0.86 0.17* −0.11* 0.31* 0.18* −0.11 7 Ln_inc 11.04 0.30 −0.11 0.12* 0.06 −0.14* 0.03 −0.1 8 Pc_osr 5.74 3.88 0.02 0.01 0.07 0.02 0.05 0.06 0.1 9 ICLEI 0.29 0.45 0.17* −0.02 0.29* 0.03 −0.07 0.32* 0.01 0.07 10 Age_hhi 0.37 0.03 0.05 0.14* 0.03 0.08 0.09 −0.11* 0.31* 0.05 −0.05 11 Eth_hhi 0.53 0.15 −0.13* 0.09 −0.17* −0.1 0.12* −0.3* 0.12* −0.06 −0.12* −0.12* 12 Pop_ch 27.88 229.64 −0.05 0.06 −0.12* −0.02 0.09 −0.04 0.03 −0.01 −0.04 0.09 −0.05 . Variable . Mean . SD . 1 . 2 . 3 . 4 . 5 . 6 . 7 . 8 . 9 . 10 . 11 . 1 Commit 0.66 0.8 2 EO 3.19 1.01 0.13* 3 Perform 2.97 0.6 0.32* 0.18* 4 Satisfied 6.95 2.49 0.15* −0.02 0.14* 5 Capacity 4.14 1.09 −0.07 0.22* −0.02 0.15* 6 Ln_pop 11.27 0.86 0.17* −0.11* 0.31* 0.18* −0.11 7 Ln_inc 11.04 0.30 −0.11 0.12* 0.06 −0.14* 0.03 −0.1 8 Pc_osr 5.74 3.88 0.02 0.01 0.07 0.02 0.05 0.06 0.1 9 ICLEI 0.29 0.45 0.17* −0.02 0.29* 0.03 −0.07 0.32* 0.01 0.07 10 Age_hhi 0.37 0.03 0.05 0.14* 0.03 0.08 0.09 −0.11* 0.31* 0.05 −0.05 11 Eth_hhi 0.53 0.15 −0.13* 0.09 −0.17* −0.1 0.12* −0.3* 0.12* −0.06 −0.12* −0.12* 12 Pop_ch 27.88 229.64 −0.05 0.06 −0.12* −0.02 0.09 −0.04 0.03 −0.01 −0.04 0.09 −0.05 Means and SD were calculated by creating averaged scores across the items that comprised the constructs, although such scales were not used in the SEM analysis and are displayed for illustration purposes only. *p < .05. Open in new tab Structural equation modeling (SEM) is appropriate to test relationships between latent constructs of EO, program performance, and program satisfaction and their manifest indicators in T1, and the commitment in T2 to continue sustainability programs. SEM offers advantages relevant to this study, including enhanced capabilities for analyzing categorical and non-normal data, as well as alternatives for dealing with missing data (Tomarken and Waller 2005). To test the first two hypotheses (H1 and H2), we applied a Bayesian structural equation modeling (BSEM) approach that offers advantages for estimating structural models with small sample sizes (Muthén and Asparouhov 2012). Recent comparisons of BSEM and SEM in the management (Iwamoto and Suzuki 2019) and tourism (Assaf, Tsionas, and Oh 2018) research suggest BSEM offers advantages over traditional covariance-based approaches when handling ordinal scale variables, which is the case for our observational data. Following Asparouhov and Muthén (2019), we estimated Bayesian latent variable interaction models using the XWITH statement in Mplus 8 (Muthén and Muthén 2012). Since we did not have a priori evidence to inform our model parameters, we used noninformative priors, which are the default in Mplus (Muthén and Asparouhov 2012). In BSEM, Markov chain Monte Carlo (MCMC) is used to estimate exact posterior distributions for model parameters and latent variables. While a posterior predictive p-value (PPP) of model fit can be obtained via a χ 2 test in noninteraction BSEM, model fit indicators are not currently available in the Bayesian latent variable interaction framework (Asparouhov and Muthén 2019). Thus, we performed an alternative latent moderated structural equation analysis, described in Supplementary Appendix B. Individual (Micro) Analytic Methods Due to the organizational-level and observational nature of these data, we cannot make causal claims about the underlying attitudes, beliefs, or behavior of public managers. Thus, the individual-level analysis utilizes a survey experiment conducted through CloudResearch, formerly known as TurkPrime, an online platform that empanels behavioral research participants. Use of online labor pools for social experiments has become a burgeoning method for psychology, political science, and public administration research, and offers some advantages over traditional lab experiments in which participants are often college students (Berinsky, Huber, and Lenz 2012; Buhrmester, Kwang, and Gosling 2011; Clifford, Jewell, and Waggoner 2015). Our panel was assembled via TurkPrime’s “concierge” service, which facilitates recruitment of specialized panels. In our case, these were US local government employees, who were asked a battery of socio-demographic questions (gender, age, income), and screened to achieve gender balance and a range of nonsupervisory, middle-management, and senior positions. The panel (N = 543) was recruited in April 2019, and individual respondents were paid $13 for completing the survey. The experiment was preregistered, and the data are available from the authors upon request. Descriptively, the panel was 72.2% white, had a modal income of over $105,000, and a modal age range of between 35 and 54 years old. In terms of seniority, 59.3% of the panel reported they did not supervise other employees, 33.2% reported they “hold a mid-level management position and directly supervise other employees,” while 7.5% reported they were a senior executive/manager. Table 2 reports the descriptive statistics for the full panel. Table 2. Descriptive Statistics for Local Government Employee Experimental Panel Variable . N . Mean . SD . Min. . Max. . Gender (proportion female) 543 0.513 0.5 0 1 Seniority (1 = no supervisory; 2 = mid-level; 3 = senior supervisor) 509 1.48 0.63 1 3 Income (modal category: 7 = $105,000 or more) 543 4.38 1.9 1 7 Age (two modal categories: 3 = 35–44 years old; 4 = 45–54 years old) 543 3.44 1.19 1 6 % White 543 0.722 0.448 0 1 Variable . N . Mean . SD . Min. . Max. . Gender (proportion female) 543 0.513 0.5 0 1 Seniority (1 = no supervisory; 2 = mid-level; 3 = senior supervisor) 509 1.48 0.63 1 3 Income (modal category: 7 = $105,000 or more) 543 4.38 1.9 1 7 Age (two modal categories: 3 = 35–44 years old; 4 = 45–54 years old) 543 3.44 1.19 1 6 % White 543 0.722 0.448 0 1 Open in new tab Table 2. Descriptive Statistics for Local Government Employee Experimental Panel Variable . N . Mean . SD . Min. . Max. . Gender (proportion female) 543 0.513 0.5 0 1 Seniority (1 = no supervisory; 2 = mid-level; 3 = senior supervisor) 509 1.48 0.63 1 3 Income (modal category: 7 = $105,000 or more) 543 4.38 1.9 1 7 Age (two modal categories: 3 = 35–44 years old; 4 = 45–54 years old) 543 3.44 1.19 1 6 % White 543 0.722 0.448 0 1 Variable . N . Mean . SD . Min. . Max. . Gender (proportion female) 543 0.513 0.5 0 1 Seniority (1 = no supervisory; 2 = mid-level; 3 = senior supervisor) 509 1.48 0.63 1 3 Income (modal category: 7 = $105,000 or more) 543 4.38 1.9 1 7 Age (two modal categories: 3 = 35–44 years old; 4 = 45–54 years old) 543 3.44 1.19 1 6 % White 543 0.722 0.448 0 1 Open in new tab In the equivalency framing experiment, local government employees were presented with a hypothetical scenario in which their organization was considering installing a methane capture and reuse system at its local wastewater treatment plant (full text of the treatment is included in Supplementary Appendix C). We then adapted a similar equivalency framing approach used in the widely replicated Asian disease outbreak to frame the choices in terms of equivalent losses or gains (Kahneman and Tversky 1986; Nicholson-Crotty, Nicholson-Crotty, and Webeck 2019). To test our hypotheses of risk-seeking and risk-aversion behavior (H3), and how EO shapes such behavior (H4), half the respondents were randomly assigned to an equivalent gains scenario and asked to choose between the following two options: Option 1 If Project A is adopted, 20% of current local government energy costs will be saved. Option 2 If Project B is adopted, there is a 1/3 probability that 60% of energy costs will be saved, and a 2/3 probability that no energy will be saved. Option 1 If Project A is adopted, 20% of current local government energy costs will be saved. Option 2 If Project B is adopted, there is a 1/3 probability that 60% of energy costs will be saved, and a 2/3 probability that no energy will be saved. Open in new tab Option 1 If Project A is adopted, 20% of current local government energy costs will be saved. Option 2 If Project B is adopted, there is a 1/3 probability that 60% of energy costs will be saved, and a 2/3 probability that no energy will be saved. Option 1 If Project A is adopted, 20% of current local government energy costs will be saved. Option 2 If Project B is adopted, there is a 1/3 probability that 60% of energy costs will be saved, and a 2/3 probability that no energy will be saved. Open in new tab The other half were assigned to an equivalent losses scenario and presented with these options: Option 1 If Project A is adopted, plant operational costs will increase 20%. Option 2 If Project B is adopted, there is a 1/3 probability that operational costs will increase 60%, and a 2/3 probability that there will be no increase in operational costs. Option 1 If Project A is adopted, plant operational costs will increase 20%. Option 2 If Project B is adopted, there is a 1/3 probability that operational costs will increase 60%, and a 2/3 probability that there will be no increase in operational costs. Open in new tab Option 1 If Project A is adopted, plant operational costs will increase 20%. Option 2 If Project B is adopted, there is a 1/3 probability that operational costs will increase 60%, and a 2/3 probability that there will be no increase in operational costs. Option 1 If Project A is adopted, plant operational costs will increase 20%. Option 2 If Project B is adopted, there is a 1/3 probability that operational costs will increase 60%, and a 2/3 probability that there will be no increase in operational costs. Open in new tab Consistent with prospect theory and our research hypotheses, we expect local government employees to overwhelmingly seek risk when costs and benefits of sustainability programs are framed in terms of losses rather than gains. If more entrepreneurial organizations are more careful risk-assessors, EO should induce risk-seeking behavior irrespective of loss or gain frames of sustainability efforts. We use Bayesian logistic regression to test our experimental hypotheses (H3 and H4) in order to make more intuitive, probabilistic statements about performance and EO. Results A central contention of this article is that because organizations face limitations on attention and routinize their response to new stimuli, they are predisposed to preserve programmatic activities. In this specific context, loss aversion plays an important role in influencing local government sustainability continuation and innovation. Thus, individuals and organizations are more likely to take risks on sustainability activities in order to avoid losses as opposed to achieving gains. We find support for our overall contention that human and organizational behavior demonstrates loss aversion, both when they consider continuing existing programs or the creation of new ones. However, contrary to government reform orthodoxy, we find mixed support for the contention that organization-level EO consistently influences this tendency. Below, we report results from each of these analyses separately, then discuss their implications in the conclusion. Organizational-Level Analysis Before testing the Bayesian latent variable interaction model, we tested a baseline BSEM model (N = 338) with no interaction terms and no control variables. The model was repeated for 40,000 samples with a potential scale reduction (PSR) of 1.06, which is below the recommended PSR cutoff value of 1.1 (Gelman et al. 2013). However, the PPP of .001 for the χ 2 test indicated a poor model fit, as values equal to .5 suggest an excellent fit since the observed data no more or less probable than the generated data (Muthén and Asparouhov 2012). Adding control variables to the model (N = 300) yielded a PSR of 1.105 after more than 43,000 iterations and a PPP of less than .001, indicating a poor fit. Nonetheless, the latent variable interaction EO × satisfied yielded a significant effect established by the 95% credible interval, which validates its inclusion in the model (Asparouhov and Muthén 2019). Table 3 reports the BSEM path values with the 95% credible intervals (Bayes factor loadings are reported in Supplementary Appendix A). Because organizational EO has been normatively hailed by government reformers as a solution to the perceptions of inept, overly bureaucratic decision making in public organizations, we tested both the direct and interaction effect of EO and satisfaction on commitment to EECBG program continuation. Bayesian models are typically interpreted with credible intervals and probabilistic statements which summarize the proportion of the posterior distribution which falls on either side of some constant, such as 0. The BSEM model suggests EO has a direct positive effect on such commitment, with 99% of the posterior density of the path value between EO and commit falling in the positive region of support. The direct effect of satisfaction on commitment to program continuation does, however, include 0 in the 95% credible interval as estimated mean of .028 falls near the center of the posterior distribution, which runs counter to H1. But, the model finds partial support for H1 in that a greater EO strengthens the relationship between satisfaction with EECBG programs and the commitment to continue them, as the path value for the interaction EO × satisfied has a greater than 99% chance of having a positive effect. Table 3. Path Values for Bayesian Latent Variable Interaction Model (N = 300) . Path . . −2.5% . µ β . 2.5% . One-tailed p-value . EO → Commit 0.031 0.17 0.310 .01 Satisfied → Commit −0.113 0.028 0.173 .351 Perform → Commit 0.053 0.203 0.345 .004 EO × satisfied → Commit 0.081 0.208 0.344 .001 EO × perform → Commit −0.272 −0.128 0.023 .043 Controls  Capacity → Commit −0.229 −0.085 0.065 .129  Ln_pop → Commit 0.058 0.171 0.284 .003  Ln_inc → Commit −0.157 −0.093 −0.026 .003  Pc_osr → Commit −0.138 −0.026 0.092 .327  ICLEI → Commit −0.048 0.07 0.188 .122  Age_hhi → Commit −0.104 0.022 0.151 .374  Eth_hhi → Commit −0.108 0.017 0.135 .394  Pop_ch → Commit −0.114 −0.006 0.105 .459 . Path . . −2.5% . µ β . 2.5% . One-tailed p-value . EO → Commit 0.031 0.17 0.310 .01 Satisfied → Commit −0.113 0.028 0.173 .351 Perform → Commit 0.053 0.203 0.345 .004 EO × satisfied → Commit 0.081 0.208 0.344 .001 EO × perform → Commit −0.272 −0.128 0.023 .043 Controls  Capacity → Commit −0.229 −0.085 0.065 .129  Ln_pop → Commit 0.058 0.171 0.284 .003  Ln_inc → Commit −0.157 −0.093 −0.026 .003  Pc_osr → Commit −0.138 −0.026 0.092 .327  ICLEI → Commit −0.048 0.07 0.188 .122  Age_hhi → Commit −0.104 0.022 0.151 .374  Eth_hhi → Commit −0.108 0.017 0.135 .394  Pop_ch → Commit −0.114 −0.006 0.105 .459 Open in new tab Table 3. Path Values for Bayesian Latent Variable Interaction Model (N = 300) . Path . . −2.5% . µ β . 2.5% . One-tailed p-value . EO → Commit 0.031 0.17 0.310 .01 Satisfied → Commit −0.113 0.028 0.173 .351 Perform → Commit 0.053 0.203 0.345 .004 EO × satisfied → Commit 0.081 0.208 0.344 .001 EO × perform → Commit −0.272 −0.128 0.023 .043 Controls  Capacity → Commit −0.229 −0.085 0.065 .129  Ln_pop → Commit 0.058 0.171 0.284 .003  Ln_inc → Commit −0.157 −0.093 −0.026 .003  Pc_osr → Commit −0.138 −0.026 0.092 .327  ICLEI → Commit −0.048 0.07 0.188 .122  Age_hhi → Commit −0.104 0.022 0.151 .374  Eth_hhi → Commit −0.108 0.017 0.135 .394  Pop_ch → Commit −0.114 −0.006 0.105 .459 . Path . . −2.5% . µ β . 2.5% . One-tailed p-value . EO → Commit 0.031 0.17 0.310 .01 Satisfied → Commit −0.113 0.028 0.173 .351 Perform → Commit 0.053 0.203 0.345 .004 EO × satisfied → Commit 0.081 0.208 0.344 .001 EO × perform → Commit −0.272 −0.128 0.023 .043 Controls  Capacity → Commit −0.229 −0.085 0.065 .129  Ln_pop → Commit 0.058 0.171 0.284 .003  Ln_inc → Commit −0.157 −0.093 −0.026 .003  Pc_osr → Commit −0.138 −0.026 0.092 .327  ICLEI → Commit −0.048 0.07 0.188 .122  Age_hhi → Commit −0.104 0.022 0.151 .374  Eth_hhi → Commit −0.108 0.017 0.135 .394  Pop_ch → Commit −0.114 −0.006 0.105 .459 Open in new tab We used the “rhdf5” package in R to create loop plots from the BSEM model estimated in Mplus. Figure 1 shows the estimated effect of EECBG program satisfaction on commitment to program continuation at high (1 SD above the 0 mean) and low EO (1 SD below the 0 mean), plotted over the program satisfaction range of −3 SD to +3 SD. In other words, when EO is high (or 1 SD above the zero mean), the estimated effect of satisfaction on commitment is positive; when EO is low (or 1 SD below the zero mean), the effect becomes negative. The 95% credible intervals for these effects are also provided. Figure 1. Open in new tabDownload slide The plot displays the moderation effect of EO on the estimated effect of satisfaction with the EECBG program in T1 on commitment to program continuation in T2, plotted over the satisfaction range of −3 SD to +3 SD. Higher EO levels induce a stronger, positive effect for satisfaction. Figure 1. Open in new tabDownload slide The plot displays the moderation effect of EO on the estimated effect of satisfaction with the EECBG program in T1 on commitment to program continuation in T2, plotted over the satisfaction range of −3 SD to +3 SD. Higher EO levels induce a stronger, positive effect for satisfaction. Turning to performance, we find support for H2, which posited perceptions of organization-wide sustainability performance have a positive, direct effect on commitment to EECBG program continuation. The BSEM model supports this claim as the path value between perform and commit has a greater than 99% chance of being positive. This may imply that when local government officials are deciding whether to continue programs, they place a greater weight on information suggesting overall sustainability performance was higher than on whether they were more satisfied with the program itself. The model suggests this occurs even when controlling for population, median household income, and ICLEI membership. The effect of population has a greater than 99% chance of being positive, while the effect of income has a greater than 99% chance of being negative, and the effect of being an ICLEI member has about an 88% chance of positively relating to commitment to program continuation. We found some support for the positive effect of EO weakening at high levels of performance (H2). The path value for the interaction EO × perform, while including 0 in the 95% credible interval, has over a 95% chance of observing a negative effect, as most of the density in the posterior distribution falls in the negative region. Local governments with higher perceived performance appear to have diminished the effect of an organizational EO on commitment to program continuation. This moderation effect, which follows the same logic as the one described above, is displayed in Figure 2. Figure 2. Open in new tabDownload slide The plot displays the moderation effect of EO on the estimated effect of perceived performance in T1 on commitment to program continuation in T2, plotted over the perceived performance range of −3 SD to +3 SD. Higher EO levels induce a weaker, negative effect for perceived performance. Figure 2. Open in new tabDownload slide The plot displays the moderation effect of EO on the estimated effect of perceived performance in T1 on commitment to program continuation in T2, plotted over the perceived performance range of −3 SD to +3 SD. Higher EO levels induce a weaker, negative effect for perceived performance. Individual-Level Analysis The second experimental design tested the effect of framing a hypothetical methane capture and reuse program in terms of equivalent gains or losses. In the group that received the gains frame, 76.5% of respondents selected the risk-averse option (Figure 3a). This result is in line with the voluminous empirical evidence supporting prospect theory (Bellé, Cantarelli, and Belardinelli 2018; Kahneman and Tversky 2013). In the loss-frame group, that risk-averse selection dropped to 56.1% (Figure 3b), which is a less-precipitous reversal than that found in recent behavioral research (Nicholson-Crotty, Nicholson-Crotty, and Webeck 2019). We first use a χ 2 test to detect whether a statistically significant relationship exists between risk tolerance and framing. The results support our third hypothesis (H3) that public employees will be more risk-seeking when the cost–benefit balance for a sustainability program is framed in terms of losses (χ 2 = 22.01, p = .000). This finding has potential normative implications considering that sustainability policy entrepreneurs and advocacy organizations tend to promote the benefits of sustainability rather than the potential economic, environmental, or social losses associated with inaction. We discuss this in greater detail in the following section. Figure 3. Open in new tabDownload slide Equivalent gains (left) and losses (right) frames in the methane program experiment demonstrate loss aversion among local administrators. The difference is less pronounced than in many replications of this experiment (Nicholson-Crotty, Nicholson-Crotty, and Webeck 2019), but the difference is statistically significant. Figure 3. Open in new tabDownload slide Equivalent gains (left) and losses (right) frames in the methane program experiment demonstrate loss aversion among local administrators. The difference is less pronounced than in many replications of this experiment (Nicholson-Crotty, Nicholson-Crotty, and Webeck 2019), but the difference is statistically significant. We then estimated two Bayesian logistic regression models with noninformative priors to examine risk aversion. The models estimate the influence of EO, past performance, and the seniority of respondents on selecting the risk-averse methane option across both gains and loss frames. To aid in model convergence, the estimation procedure involved sampling 240,000 pulls (with a 40,000 “burn-in” which is discarded) from a simulated posterior distribution via MCMC approach using the Metropolis–Hastings algorithm (Gill and Witko 2013). Table 4 reports the computed posterior means, Monte Carlo standard errors (MCSE) and credible intervals. Table 4. Bayesian Logistic Regressions for Methane Program Gains and Losses Framing . Risk Aversion with Gains Frame . Risk Aversion with Losses Frame . . Mean . MCSE . Credible Interval . Mean . MCSE . Credible Interval . Performance 6.13 0.019 1.04; 13.3 −1.19 0.009 −3.98; 1.55 EO .862 0.024 −3.62; 5.72 1.8 0.024 −2.19; 5.92 Seniority −0.179 0.004 −.674; 0.32 −0.528 0.004 −0.999; −0.066 Female −0.268 0.004 −0.937; 0.396 −0.546 0.003 −1.09; −0.012 Income −0.147 0.001 −0.326; 0.033 0.023 0.001 −0.125; 0.167 Age 0.279 0.003 −0.032; 0.599 0.071 0.002 −0.144; 0.288 White 0.803 0.003 0.122; 1.49 0.036 0.003 −0.556; 0.628 MCMC 240,000 240,000 Obs 244 246 Acceptance 0.442 0.454 Efficiency 0.044 0.039 . Risk Aversion with Gains Frame . Risk Aversion with Losses Frame . . Mean . MCSE . Credible Interval . Mean . MCSE . Credible Interval . Performance 6.13 0.019 1.04; 13.3 −1.19 0.009 −3.98; 1.55 EO .862 0.024 −3.62; 5.72 1.8 0.024 −2.19; 5.92 Seniority −0.179 0.004 −.674; 0.32 −0.528 0.004 −0.999; −0.066 Female −0.268 0.004 −0.937; 0.396 −0.546 0.003 −1.09; −0.012 Income −0.147 0.001 −0.326; 0.033 0.023 0.001 −0.125; 0.167 Age 0.279 0.003 −0.032; 0.599 0.071 0.002 −0.144; 0.288 White 0.803 0.003 0.122; 1.49 0.036 0.003 −0.556; 0.628 MCMC 240,000 240,000 Obs 244 246 Acceptance 0.442 0.454 Efficiency 0.044 0.039 Diagnostic plots present strong evidence of model convergence. MCSE, Monte Carlo standard errors. Open in new tab Table 4. Bayesian Logistic Regressions for Methane Program Gains and Losses Framing . Risk Aversion with Gains Frame . Risk Aversion with Losses Frame . . Mean . MCSE . Credible Interval . Mean . MCSE . Credible Interval . Performance 6.13 0.019 1.04; 13.3 −1.19 0.009 −3.98; 1.55 EO .862 0.024 −3.62; 5.72 1.8 0.024 −2.19; 5.92 Seniority −0.179 0.004 −.674; 0.32 −0.528 0.004 −0.999; −0.066 Female −0.268 0.004 −0.937; 0.396 −0.546 0.003 −1.09; −0.012 Income −0.147 0.001 −0.326; 0.033 0.023 0.001 −0.125; 0.167 Age 0.279 0.003 −0.032; 0.599 0.071 0.002 −0.144; 0.288 White 0.803 0.003 0.122; 1.49 0.036 0.003 −0.556; 0.628 MCMC 240,000 240,000 Obs 244 246 Acceptance 0.442 0.454 Efficiency 0.044 0.039 . Risk Aversion with Gains Frame . Risk Aversion with Losses Frame . . Mean . MCSE . Credible Interval . Mean . MCSE . Credible Interval . Performance 6.13 0.019 1.04; 13.3 −1.19 0.009 −3.98; 1.55 EO .862 0.024 −3.62; 5.72 1.8 0.024 −2.19; 5.92 Seniority −0.179 0.004 −.674; 0.32 −0.528 0.004 −0.999; −0.066 Female −0.268 0.004 −0.937; 0.396 −0.546 0.003 −1.09; −0.012 Income −0.147 0.001 −0.326; 0.033 0.023 0.001 −0.125; 0.167 Age 0.279 0.003 −0.032; 0.599 0.071 0.002 −0.144; 0.288 White 0.803 0.003 0.122; 1.49 0.036 0.003 −0.556; 0.628 MCMC 240,000 240,000 Obs 244 246 Acceptance 0.442 0.454 Efficiency 0.044 0.039 Diagnostic plots present strong evidence of model convergence. MCSE, Monte Carlo standard errors. Open in new tab We then conducted Bayesian interval hypothesis tests, reported in Table 5, to estimate the probabilities that each of these parameters positively influence the risk-averse choice. We find a 99.4% chance that perceptions of local government sustainability performance are positively related to risk aversion in the gains-framing scenario. This result is consistent with the EECBG findings; past performance appears to drive risk aversion when the scenario is presented as a potential gain. However, when the program is framed in terms of potential losses, we find only a 19.5% chance of observing this performance effect. We conclude that these findings offer evidence that past performance works as a heuristic for assessments of the benefits of future program participation, consistent with our observational results. Turning to EO, we find only a 63.4% chance that EO is positively related to risk aversion in the gains scenario and an 81% chance in the loss scenario, which we consider inconclusive evidence for H4. Finally, we observe a 23.7% chance that more senior supervisors are risk averse under the gains scenario, but only a 1.3% chance in the loss frame. In other words, we find a 98.7% chance that seniority in an organizational hierarchy positively influences risk-taking in the loss scenario. This suggests that, relative to subordinates, higher-level administrators in local governments are more comfortable taking risks under a loss scenario. This could be because they conceivably have more to lose (such as their jobs) from poor performance or the likelihood that more ambitious, risk-taking individuals self-select into the higher rungs of public organizations. In either case, this finding speaks to the importance of a micro-level approach for organizational studies. While lower-level buy-in, delegation and participatory decision-making are hailed as important for successful performance management in local governments, higher-level administrators clearly play a central role in organization-wide decisions. Table 5. Bayesian Interval Hypothesis Tests for Risk Aversion Variable . Interval Hypotheses . Gain Frame . Loss Frame . Performance p > 0 .994 .195 EO p > 0 .634 .81 Seniority p > 0 .237 .013 Variable . Interval Hypotheses . Gain Frame . Loss Frame . Performance p > 0 .994 .195 EO p > 0 .634 .81 Seniority p > 0 .237 .013 Open in new tab Table 5. Bayesian Interval Hypothesis Tests for Risk Aversion Variable . Interval Hypotheses . Gain Frame . Loss Frame . Performance p > 0 .994 .195 EO p > 0 .634 .81 Seniority p > 0 .237 .013 Variable . Interval Hypotheses . Gain Frame . Loss Frame . Performance p > 0 .994 .195 EO p > 0 .634 .81 Seniority p > 0 .237 .013 Open in new tab Discussion and Conclusion A primary logic of NPM and other government reforms is that the tendencies of Homo bureaucraticus can be overcome through indoctrinating or designing entrepreneurialism into public organizations. Loss aversion represents an important cognitive caveat to this creed. EO and the broader notion of entrepreneurialism was imported from the generic management literature (Lumpkin and Dess 1996) and implies that for public organizations to be more risk-taking, innovative, and proactive, they must be willing to accept some losses to achieve overarching goals or greater gains (Meynhardt and Diefenbach 2012; Teske and Schneider 1994). But evidence from prospect theory and its underlying principle of bounded rationality has conclusively demonstrated that loss aversion can bias how individuals contemplate risks versus rewards (Bellé, Cantarelli, and Belardinelli 2018; Kahneman and Tversky 2013). A behavioral theory of collective choice applied to public organizations holds that government officials may be more likely to take risks with public resources when they are confronted with potential losses. The evidence that local governments discounted their direct experiences with EECBG-funded activities when deciding to continue them is consistent with March and Simon’s (1958) hypothesis that innovation must overcome the sunk costs which work against abandoning existing programs. In scenarios where local government managers may have more time to engage in “double-loop” learning and adapting, it is possible that direct experience with the program may play a larger role in assessing risk. However, our study suggests local governments in this instance defaulted to another heuristic: their diffuse, broader experience with sustainability efforts. Experimental results extend this managerial loss aversion finding to equivalent loss and gains scenarios at the micro-level. We find that higher organizational performance increases risk aversion at the individual level, although senior administrators were more likely to be risk-taking. A logical next step for this research is to examine how individual biases are aggregated into organizational decisions or the micro- to meso-level mechanisms. This likely requires field experimentation in which micro-level treatments and resultant organization decisions can be observed. At a minimum, the results from this study suggest that reforms designed to engender an organizational EO may not be a panacea for sustainability challenges facing governments globally. Our study raises practical considerations for how government officials frame the challenges of sustainability for their communities, particularly now that local governments have been significantly affected by the COVID-19 pandemic. Organizations such as ICLEI-Local Governments for Sustainability, the US Conference of Mayors, the International City/County Management Association, the Global Covenant of Mayors for Climate & Energy, and others have spent years, if not decades, organizing and encouraging local governments to take voluntary actions to improve sustainability (Bai et al. 2018; Homsy and Warner 2015; Krause 2011; Portney 2013). Such choices are typically framed around energy efficiency, quality-of-life and other cobenefits, or global environmental gains, which could be achieved from investments in sustainability, and criticized for the potential economic or property-rights losses such commitments could impose (Berry and Portney 2017; Krause, Yi, and Feiock 2016; Yi, Krause, and Feiock 2017). Barriers typically identified include a lack of public engagement, a focus on short-term savings over longer-term benefits, and prohibitive intergovernmental and intersectoral collaboration costs (Deslatte and Feiock 2019; Portney 2005; Zeemering 2014). However, if local governments are more likely to take risks when facing potential losses, this suggests that sustainability advocacy efforts premised on touting potential societal gains may be wasting an opportunity. Framing sustainability in terms of potential social, economic, and environmental losses, which may result from inaction, has the potential to motivate greater support for such initiatives. To some extent, this framing is already occurring as the scientific community and media focus communications strategies more on the negative impacts of climate change. For instance, the May 2019 report from the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services1 drew worldwide media and popular attention by highlighting the anthropogenic-induced increase in extinction risks for one million species. Organizations such as the Union of Concerned Scientists,2 universities,3 and the media4 have increased public and stakeholder outreach efforts to link extreme precipitation, flooding, and climate change. In a noteworthy style-change, The Guardian newspaper announced in May 2019 that it would begin using phrases such as “climate crisis” and “global heating” to emphasis the dire nature of the threat, prompting other news organizations to consider following suit.5 This indicates that public administration scholars need to direct greater attention to the risk-taking tendencies of local government officials as future infrastructure, ecosystem, social, and economic consequences of climate change are made more salient. This trend also highlights a messaging strategy that goes beyond equivalence framing. Issue framing, in which substantively distinct considerations surrounding a policy issue are raised and acted upon, is more commonly used in political science to explore the effects of campaign messaging, elite mobilization, and party competition (Druckman 2004; Slothuus and de Vreese 2010). For instance, one such study on the influence of competing issue frames surrounding a local government growth management initiative—that it would preserve green space but impose economic costs—found that citizen exposure to either the strong “pro” or “con” frames via editorials significantly swayed support above or below the mean, respectively (Chong and Druckman 2007). No studies of which we are aware explore the influence of issue framing on public managers’ decision making (Battaglio et al. 2018; Deslatte 2020b; Laurian, Walker, and Crawford 2017). This is a surprising omission considering the extent to which local managers are immersed in substantive issue-based arguments for or against the myriad social, environmental, and economic policy problems cities confront (Deslatte, Swann, and Feiock 2017). Future research should direct attention to this omission and examine the effects of positive and negative issue framing on organizational identification, public and stakeholder support. The limited evidence we find supporting a role for EO in sustainability initiatives also has implications for policy design. Our analysis found an EO only moderates the influence of satisfaction on sustainability program continuation and plays little direct role in decisions to launch new programs. One implication from loss aversion is that local governments may be more likely to make investments to preserve sustainability programs initiated through state or federal initiatives as opposed to instigating them on their own. As previously discussed, this is because the limits on the speed and collective cognitive capacities of organizations make preserving existing programmatic activities more appealing than directing activities toward the creation of new programs (March and Simon 1958). The decentralized characteristics of US fiscal federalism make programs such as EECBG appealing mechanisms from a policy design perspective for stimulating innovation and state and local compliance with national policy goals (Terman 2015; Terman et al. 2016; Wang and Pagano 2017). However, the broader federal stimulus act which authorized EECBG funding was passed by Congress largely along party lines, encountered accountability and transparency challenges (Hall and Jennings 2011; Jennings, Hall, and Zhang 2012), and significant variation across states in implementation timelines and successes (Carley, Nicholson-Crotty, and Fisher 2015). While local-level climate action has been touted by advocacy groups as a remedy to the US decision to leave the 2015 Paris climate accord, loss aversion presents a potential barrier to notions that substate actors can go it alone. If more entrepreneurially oriented local governments are no more likely to initiate major strides toward sustainability, local government voluntary actions may not be enough (Portney 2013; Swann 2017). Given the scale and immediacy of climate change threats, the architecture of organizational decision making suggests future federal initiation of sustainability efforts will be necessary, in much the same way that traditional Community Development Block Grants have evolved into an ongoing intergovernmental program. While there remain significant gaps in our understanding of all that motivates organizational change, the theoretical components of bounded rationality have provided tools for building better institutions to support such future efforts. Footnotes 1 https://www.ipbes.net/sites/default/files/downloads/spm_unedited_advance_for_posting_htn.pdf. 2 https://www.ucsusa.org/global-warming/global-warming-impacts/floods. 3 https://eri.iu.edu/prepare/webinars.html. 4 https://www.wired.com/story/for-the-midwest-epic-flooding-is-the-face-of-climate-change/. 5 https://www.theguardian.com/environment/2019/may/24/media-outlets-guardian-reconsider-language-climate. References Anechiarico , Frank . 2007 . The new public management at middle age: Critiques of the performance movement . Public Administration Review 67 ( 4 ): 783 – 86 . OpenURL Placeholder Text WorldCat Asparouhov , Tihomir , and Bengt Muthén. 2019 . Bayesian Estimation of Single and Multilevel Models with Latent Variable Interactions. http://www.statmodel.com/download/LVinteractions.pdf (accessed October 20, 2019 ). Assaf , A. George , Mike Tsionas, and Haemoon Oh. 2018 . The time has come: Toward Bayesian SEM estimation in tourism research . Tourism Management 64 ( February ): 98 – 109 . OpenURL Placeholder Text WorldCat Audretsch , David B . 2015 . Everything in Its Place: Entrepreneurship and the Strategic Management of Cities, Regions, and States . New York, NY : Oxford Univ. Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Bai , Xuemei , Richard J. Dawson, Diana Ürge-Vorsatz, Gian C. Delgado, Aliyu Salisu Barau, Shobhakar Dhakal, Dodman David, et al. 2018 . Six research priorities for cities and climate change . Nature 555 ( 7694 ): 23 – 25 . OpenURL Placeholder Text WorldCat Battaglio , R. Paul , Jr., Paolo Belardinelli, Nicola Bellé, and Paola Cantarelli. 2018 . Behavioral public administration Ad Fontes: A synthesis of research on bounded rationality, cognitive biases, and nudging in public organizations . Public Administration Review 25 ( September ): 599 . OpenURL Placeholder Text WorldCat Belardinelli , Paolo , Nicola Bellé, Mariafrancesca Sicilia, and Ileana Steccolini. 2018 . Framing effects under different uses of performance information: An experimental study on public managers . Public Administration Review 78 ( 6 ): 841 – 51 . OpenURL Placeholder Text WorldCat Bellé , Nicola , Paola Cantarelli, and Paolo Belardinelli. 2018 . Prospect theory goes public: Experimental evidence on cognitive biases in public policy and management decisions . Public Administration Review 78 ( 6 ): 828 – 40 . OpenURL Placeholder Text WorldCat Bellone , Carl J. , and George Frederick Goerl. 1992 . Reconciling public entrepreneurship and democracy . Public Administration Review 52 ( 2 ): 130 – 4 . OpenURL Placeholder Text WorldCat Berinsky , Adam J. , Gregory A. Huber, and Gabriel S. Lenz. 2012 . Evaluating online labor markets for experimental research: Amazon.com’s Mechanical Turk . Political Analysis 20 ( 3 ): 351 – 68 . OpenURL Placeholder Text WorldCat Berry , William D. , Matt Golder, and Daniel Milton. 2012 . Improving tests of theories positing interaction . The Journal of Politics 74 ( 3 ): 653 – 71 . OpenURL Placeholder Text WorldCat Berry , Jeffrey M. , and Kent E. Portney. 2017 . The Tea Party versus Agenda 21: Local groups and sustainability policies in U.S. cities . Environmental Politics 26 ( 1 ): 118 – 37 . OpenURL Placeholder Text WorldCat Brinton Milward , H. , and Hal G. Rainey. 1983 . Don’t blame the bureaucracy! Journal of Public Policy 3 ( 2 ): 149 – 68 . OpenURL Placeholder Text WorldCat Buhrmester , Michael , Tracy Kwang, and Samuel D. Gosling. 2011 . Amazon’s Mechanical Turk: A new source of inexpensive, Yet High-Quality, Data? Perspectives on Psychological Science 6 ( 1 ): 3 – 5 . OpenURL Placeholder Text WorldCat Bullock , Justin B. , Robert A. Greer, and Laurence J. O’Toole. 2019 . Managing risks in public organizations: A conceptual foundation and research agenda . Perspectives on Public Management and Governance 2 ( 1 ): 75 – 87 . OpenURL Placeholder Text WorldCat Carley , Sanya , Sean Nicholson-Crotty, and Eric J. Fisher. 2015 . Capacity, guidance, and the implementation of the American recovery and reinvestment act . Public Administration Review 75 ( 1 ): 113 – 25 . OpenURL Placeholder Text WorldCat Carr , Jered B . 2015 . What have we learned about the performance of council-manager government? A review and synthesis of the research . Public Administration Review 75 ( 5 ): 673 – 89 . OpenURL Placeholder Text WorldCat Chong , Dennis , and James N. Druckman. 2007 . Framing public opinion in competitive democracies . The American Political Science Review 101 ( 4 ): 637 – 55 . OpenURL Placeholder Text WorldCat Clifford , Scott , Ryan M. Jewell, and Philip D. Waggoner. 2015 . Are samples drawn from mechanical turk valid for research on political ideology? Research & Politics 2 ( 4 ): 2053168015622072 . OpenURL Placeholder Text WorldCat Cohen , Michael D. , James G. March, and Johan P. Olsen. 1972 . A garbage can model of organizational choice . Administrative Science Quarterly 17 ( 1 ): 1 – 25 . OpenURL Placeholder Text WorldCat Deslatte , Aaron . 2018 . Managerial friction and land-use policy punctuations in the fragmented metropolis . Policy Studies Journal doi: 10.1111/psj.12275. OpenURL Placeholder Text WorldCat ———. 2019 . A Bayesian approach for behavioral public administration: Citizen assessments of local government sustainability performance . Journal of Behavioral Public Administration 2 ( 1 ): 1 – 12 . OpenURL Placeholder Text WorldCat ———. 2020a . Positivity and negativity dominance in citizen assessments of intergovernmental sustainability performance . Journal of Public Administration Research and Theory . doi:10.1093/jopart/muaa004. Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC ———. 2020b . Revisiting bureaucratic entrepreneurialism in the age of urban austerity: Framing issues, taking risks, and building collaborative capacity . In Urban Studies and Entrepreneurship , eds. Muhammad Naveed Iftikhar, Jonathan B. Justice, and David B. Audretsch, 143 – 59 . Cham, Switzerland : Springer International Publishing . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Deslatte , Aaron , and Richard C. Feiock. 2019 . The collaboration riskscape: Fragmentation, problem types and preference divergence in urban sustainability . Publius 49 ( 2 ): 352 – 77 . OpenURL Placeholder Text WorldCat Deslatte , Aaron , and William L. Swann. 2016 . Is the price right? Gauging the marketplace for local sustainable policy tools . Journal of Urban Affairs 38 ( 4 ): 581 – 96 . OpenURL Placeholder Text WorldCat ———. 2020 . Elucidating the linkages between entrepreneurial orientation and local government sustainability performance . American Review of Public Administration 50 ( 1 ): 92 – 109 . OpenURL Placeholder Text WorldCat Deslatte , Aaron , William L. Swann, and Richard C. Feiock. 2017 . Three sides of the same coin? A Bayesian analysis of strategic management, comprehensive planning, and inclusionary values in land use . Journal of Public Administration Research and Theory 27 ( 3 ): 415 – 32 . OpenURL Placeholder Text WorldCat Dess , Gregory G. , and Donald W. Beard. 1984 . Dimensions of organizational task environments . Administrative Science Quarterly 29 ( 1 ): 52 – 73 . OpenURL Placeholder Text WorldCat DNV GL . 2015 . National evaluation of the energy efficiency and conservation block grant program, Executive SUmmary , 1 – 30 . https://weatherization.ornl.gov/wp-content/uploads/pdf/EECGBEvaluation/EECBG_ReportExecutiveSummaryFinal.pdf (Accessed June 23, 2019 ). OpenURL Placeholder Text WorldCat Deutsch , Karl W . 1965 . The nerves of government: Models of communication and control . Ethics 75 ( 4 ): 301 – 4 . OpenURL Placeholder Text WorldCat Druckman , James N . 2001 . The implications of framing effects for citizen competence . Political Behavior 23 ( 3 ): 225 – 56 . OpenURL Placeholder Text WorldCat ———. 2004 . Political preference formation: Competition, deliberation, and the (ir)relevance of framing effects . The American Political Science Review 98 ( 4 ): 671 – 86 . OpenURL Placeholder Text WorldCat Dunlop , Claire A. , and Claudio M. Radaelli. 2013 . Systematising policy learning: From monolith to dimensions . Political Studies 61 ( 3 ): 599 – 619 . OpenURL Placeholder Text WorldCat ———. 2017 . Learning in the bath-tub: The micro and macro dimensions of the causal relationship between learning and policy change . Policy and Society 36 ( 2 ): 304 – 19 . OpenURL Placeholder Text WorldCat Favero , Nathan , and Justin B. Bullock. 2015 . How (Not) to Solve the Problem: An Evaluation of Scholarly Responses to Common Source Bias . Journal of Public Administration Research and Theory 25 ( 1 ): 285 – 308 . OpenURL Placeholder Text WorldCat Feeney , Mary K . 2012 . Organizational red tape: A measurement experiment . Journal of Public Administration Research and Theory 22 ( 3 ): 427 – 44 . OpenURL Placeholder Text WorldCat Fernandes , Ronald , and Herbert A. Simon. 1999 . A study of how individuals solve complex and ill-structured problems . Policy Sciences 32 ( 3 ): 225 – 45 . OpenURL Placeholder Text WorldCat Fiorino , Daniel J . 2018 . A Good Life on a Finite Earth: The Political Economy of Green Growth . New York, NY : Oxford Univ. Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Florida , Richard , Patrick Adler, and Charlotta Mellander. 2017 . The city as innovation machine . Regional Studies 51 ( 1 ): 86 – 96 . OpenURL Placeholder Text WorldCat Funke , Joachim . 2001 . Dynamic systems as tools for analysing human judgement . Thinking & Reasoning 7 ( 1 ): 69 – 89 . OpenURL Placeholder Text WorldCat Gelman , Andrew , John B. Carlin, Hal S. Stern, David B. Dunson, Aki Vehtari, and Donald B. Rubin. 2013 . Bayesian Data Analysis . 3rd ed. Boca Raton, FL : CRC Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC George , B. , M. Baekgaard, A. Decramer, M. Audenaert, and S. Goeminne. 2020 . Institutional isomorphism, negativity bias and performance information use by politicians: A survey experiment . Public Administration 98 ( 1 ): 14 – 28 . OpenURL Placeholder Text WorldCat Gerhard , Carla , Andreas G. Klein, Karin Schermelleh-Engel, Helfried Moosbrugger, Jana Gäde, and Holger Brandt. 2015 . On the performance of likelihood-based difference tests in nonlinear structural equation models . Structural Equation Modeling: A Multidisciplinary Journal 22 ( 2 ): 276 – 87 . OpenURL Placeholder Text WorldCat Gill , Jeff , and Christopher Witko. 2013 . Bayesian analytical methods: A methodological prescription for public administration . Journal of Public Administration Research and Theory 23 ( 2 ): 457 – 94 . OpenURL Placeholder Text WorldCat Goggin , Malcolm L . 1990 . Implementation Theory and Practice: Toward a Third Generation . Glenview, IL : Scott Foresman & Co . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Grimmelikhuijsen , Stephan G. , and Albert J. Meijer. 2012 . Effects of transparency on the perceived trustworthiness of a government organization: Evidence from an online experiment . Journal of Public Administration Research and Theory 24 ( 1 ): 137 – 57 . OpenURL Placeholder Text WorldCat Grosso , Ashley , Étienne Charbonneau, and Gregg G. Van Ryzin. 2017 . How citizens respond to outputs, outcomes, and costs: A survey experiment about an HIV/AIDS program . International Public Management Journal 20 ( 1 ): 160 – 81 . OpenURL Placeholder Text WorldCat Hall , Jeremy L. , and Edward T. Jennings. 2011 . The American Recovery and Reinvestment Act (ARRA) a critical examination of accountability and transparency in the Obama administration . Public Performance & Management Review 35 ( 1 ): 202 – 26 . OpenURL Placeholder Text WorldCat Head , Brian W. , and John Alford. 2015 . Wicked Problems: Implications for public policy and management . Administration & Society 47 ( 6 ): 711 – 39 . OpenURL Placeholder Text WorldCat Heclo , Hugh . 2010 . Modern Social Politics in Britain and Sweden . Colchester, UK : ECPR Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Hendren , Kathryn , Qian Eric Luo, and Sanjay K. Pandey. 2018 . The state of mixed methods research in public administration and public policy . Public Administration Review 78 ( 6 ): 904 – 16 . OpenURL Placeholder Text WorldCat Homsy , George C. , and Mildred E. Warner. 2015 . Cities and sustainability: Polycentric action and multilevel governance . Urban Affairs Review 51 ( 1 ): 46 – 73 . OpenURL Placeholder Text WorldCat Hughes , Sara . 2019 . Repowering Cities: Governing Climate Change Mitigation in New York City, Los Angeles, and Toronto . Ithaca, NY : Cornell Univ. Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Ingraham , Patricia Wallace . 2005 . Performance: Promises to keep and miles to go . Public Administration Review 65 ( 4 ): 390 – 95 . OpenURL Placeholder Text WorldCat Iwamoto , Hiroki , and Hideo Suzuki. 2019 . An empirical study on the relationship of corporate financial performance and human capital concerning corporate social responsibility: Applying SEM and Bayesian SEM . Cogent Business & Management 6 ( 1 ): 1656443 . OpenURL Placeholder Text WorldCat James , Oliver , Sebastian R. Jilke, and Gregg G. Van Ryzin. 2017 . Experiments in Public Management Research: Challenges and Contributions . Cambridge, UK : Cambridge Univ. Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Jennings , Edward T. , Jeremy L. Hall, and Zhiwei Zhang. 2012 . The American recovery and reinvestment act and state accountability . Public Performance & Management Review 35 ( 3 ): 527 – 49 . OpenURL Placeholder Text WorldCat Jilke , Sebastian , Gregg G. Van Ryzin, and Steven Van de Walle. 2016 . Responses to decline in Marketized public services: An experimental evaluation of choice overload . Journal of Public Administration Research and Theory 26 ( 3 ): 421 – 32 . OpenURL Placeholder Text WorldCat Jones , Bryan D . 1994 . Reconceiving Decision-Making in Democratic Politics: Attention, Choice, and Public Policy . Chicago, IL : Univ. of Chicago Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC ———. 2001 . Politics and the Architecture of Choice: Bounded Rationality and Governance . Chicago, IL : Univ. of Chicago Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC ———. 2002 . Bounded rationality and public policy: Herbert A. Simon and the decisional foundation of collective choice . Policy Sciences 35 ( 3 ): 269 – 84 . OpenURL Placeholder Text WorldCat ———. 2003 . Bounded rationality and political science: Lessons from public administration and public policy . Journal of Public Administration Research and Theory 13 ( 4 ): 395 – 412 . OpenURL Placeholder Text WorldCat Jones , Bryan D. , and Frank R. Baumgartner. 2012 . From there to here: Punctuated equilibrium to the general punctuation thesis to a theory of government information processing . Policy Studies Journal 40 ( 1 ): 1 – 20 . Jones , Bryan D. , and Peter B. Mortensen. 2018 . Punctuated equilibrium theory: Explaining stability and change in public policymaking, Frank R. Baumgartner . In Theories of the Policy Process , ed. Christopher M. Weible and Paul A. Sabatier, 65 – 112 . New York, NY : Routledge . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Kahneman , Daniel . 2011 . Thinking, Fast and Slow . Vol. 1 . New York : Farrar, Straus and Giroux . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Kahneman , Daniel , and Amos Tversky. 1979 . Prospect theory: An analysis of decision under risk . Econometrica 47 ( 2 ): 263 – 91 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC ———. 1986 . Rational choice and the framing of decisions . Journal of Business 59 ( 4 ): 251 – 78 . OpenURL Placeholder Text WorldCat ———. 2013 . Prospect theory: An analysis of decision under risk. In Handbook of the Fundamentals of Financial Decision Making: Part I , ed. Leonard C. MacLean and William T. Ziemba, 99 – 127 . Singapore : World Scientific . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Kearney , Claudine , and Timo Meynhardt. 2016 . Directing corporate entrepreneurship strategy in the public sector to public value: Antecedents, components, and outcomes . International Public Management Journal 19 ( 4 ): 543 – 72 . OpenURL Placeholder Text WorldCat Kettl , Donald F . 2016 . Escaping Jurassic Government: How to Recover America’s Lost Commitment to Competence . Washington, DC : Brookings Institution Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Kingdon , John W. , and James A. Thurber. 1984 . Agendas, Alternatives, and Public Policies . Vol. 45 . Boston, MA : Little, Brown . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Klein , Andreas , and Helfried Moosbrugger. 2000 . Maximum likelihood estimation of latent interaction effects with the LMS method . Psychometrika 65 ( 4 ): 457 – 74 . OpenURL Placeholder Text WorldCat Krause , Rachel M . 2011 . Symbolic or substantive policy? Measuring the extent of local commitment to climate protection . Environment and Planning C Government & Policy 29 ( 1 ): 46 – 62 . OpenURL Placeholder Text WorldCat Krause , Rachel M. , Hongtao Yi, and Richard C. Feiock. 2016 . Applying policy termination theory to the abandonment of climate protection initiatives by U.S. local governments . Policy Studies Journal 44 ( 2 ): 176 – 95 . OpenURL Placeholder Text WorldCat Laurian , Lucie , Mark Walker, and Jan Crawford. 2017 . Implementing environmental sustainability in local government: The impacts of framing, agency culture, and structure in US cities and counties . International Journal of Public Administration 40 ( 3 ): 270 – 83 . OpenURL Placeholder Text WorldCat Lee , Danbee , and Gregg G. Van Ryzin. 2020 . Bureaucratic reputation in the eyes of citizens: An analysis of US federal agencies . International Review of Administrative Sciences 86 ( 1 ): 183 – 200 . OpenURL Placeholder Text WorldCat Lumpkin , G. T. , and Gregory G. Dess. 1996 . Clarifying the entrepreneurial orientation construct and linking it to performance . AMRO 21 ( 1 ): 135 – 72 . OpenURL Placeholder Text WorldCat March , James G. , and Herbert A. Simon 1958 . Organizations . Hoboken, NJ : Wiley . Maslowsky , Julie , Justin Jager, and Douglas Hemken. 2015 . Estimating and interpreting latent variable interactions: A tutorial for applying the latent moderated structural equations method . International Journal of Behavioral Development 39 ( 1 ): 87 – 96 . OpenURL Placeholder Text WorldCat Mazmanian , Daniel A. , and Michael E. Kraft. 2009 . Toward Sustainable Communities: Transition and Transformations in Environmental Policy . Cambridge, MA : MIT Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Mazmanian , Daniel A. , and Paul A. Sabatier. 1983 . Implementation and Public Policy. Glenview, IL : Scott Foresman . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Meier , Kenneth J. , Nathan Favero, and Ling Zhu. 2015 . Performance gaps and managerial decisions: A Bayesian decision theory of managerial action . Journal of Public Administration Research and Theory 25 ( 4 ): 1221 – 46 . OpenURL Placeholder Text WorldCat Mele , Valentina , and Paolo Belardinelli. 2019 . Mixed methods in public administration research: Selecting, sequencing, and connecting . Journal of Public Administration Research and Theory 29 ( 2 ): 334 – 47 . OpenURL Placeholder Text WorldCat Meynhardt , Timo , and Fabian E. Diefenbach. 2012 . What drives entrepreneurial orientation in the public sector? Evidence from Germany’s federal labor agency . Journal of Public Administration Research and Theory 22 ( 4 ): 761 – 92 . OpenURL Placeholder Text WorldCat Morris , Michael H. , and Foard F. Jones. 1999 . Entrepreneurship in established organizations: The case of the public sector . Entrepreneurship Theory and Practice 24 ( 1 ): 71 – 91 . OpenURL Placeholder Text WorldCat Moynihan , Donald P . 2009 . Through a glass, darkly: Understanding the effects of performance regimes . Public Performance & Management Review 32 ( 4 ): 592 – 603 . OpenURL Placeholder Text WorldCat Moyson , Stéphane , Peter Scholten, and Christopher M. Weible. 2017 . Policy learning and policy change: Theorizing their relations from different perspectives . Policy and Society 36 ( 2 ): 161 – 77 . OpenURL Placeholder Text WorldCat Muthén , Bengt , and Tihomir Asparouhov. 2012 . Bayesian structural equation modeling: A more flexible representation of substantive theory . Psychological Methods 17 ( 3 ): 313 – 35 . OpenURL Placeholder Text WorldCat Muthén , L. K. , and B. O. Muthén. 2012 . Mplus User’s Guide . 7th ed. Los Angeles, CA : Muthén & Muthén . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Nicholson-Crotty , Sean . 2004 . Goal conflict and fund diversion in federal grants to the states . American Journal of Political Science 48 ( 1 ): 110 – 22 . OpenURL Placeholder Text WorldCat Nicholson‐Crotty , Sean , Jill Nicholson‐Crotty, and Sergio Fernandez. 2017 . Performance and management in the public sector: Testing a model of relative risk aversion . Public Administration Review 77 ( 4 ): 603 – 14 . OpenURL Placeholder Text WorldCat Nicholson-Crotty , Sean , Jill Nicholson-Crotty, and Sean Webeck. 2019 . Are public managers more risk averse? Framing effects and status quo bias across the sectors . Journal of Behavioral Public Administration 2 ( 1 ). doi:10.30636/jbpa.21.35. OpenURL Placeholder Text WorldCat Olsen , Asmus Leth . 2015 . Citizen (dis)satisfaction: An experimental equivalence framing study . Public Administration Review 75 ( 3 ): 469 – 78 . OpenURL Placeholder Text WorldCat Osborne , David , and Ted Gaebler. 1992. Reinventing Government: How the Entrepreneurial Spirit is Transforming the Public Sector . New York, NY : Penguin Books . Piotrowski , Suzanne , Stephan Grimmelikhuijsen, and Felix Deat. 2019 . Numbers over narratives? How government message strategies affect citizens’ attitudes . Public Performance & Management Review 42 ( 5 ): 1005 – 28 . OpenURL Placeholder Text WorldCat Portney , Kent E. . 2005 . Civic engagement and sustainable cities in the United States . Public Administration Review 65 ( 5 ): 579 – 91 . OpenURL Placeholder Text WorldCat Portney , Kent E . 2013 . Taking Sustainable Cities Seriously: Economic Development, the Environment, and Quality of Life in American Cities . Cambridge, MA : MIT Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Pressman , Jeffrey L. , and Aaron Wildavsky. 1973 . Implementation: How Great Expectations in Washington Are Dashed in Oakland; Or, Why It’s Amazing That Federal Programs Work at All, This Being a Saga... Morals on a Foundation (Oakland Project). Berkeley, CA : Univ. of California Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Redelmeier , Donald A. , Joel Katz, and Daniel Kahneman. 2003 . Memories of colonoscopy: A randomized trial . Pain 104 ( 1–2 ): 187 – 94 . OpenURL Placeholder Text WorldCat Rutherford , Amanda , and Kenneth J. Meier. 2015 . Managerial goals in a performance-driven system: Theory and empirical tests in higher education . Public Administration 93 ( 1 ): 17 – 33 . OpenURL Placeholder Text WorldCat Simon , Herbert A . 1991 . Bounded rationality and organizational learning . Organization Science 2 ( 1 ): 125 – 34 . OpenURL Placeholder Text WorldCat ———. 1996 . The Sciences of the Artificial . Cambridge, MA : MIT Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Singla , Akheil , Justin M. Stritch, and Mary K. Feeney. 2018 . Constrained or creative? Changes in financial condition and entrepreneurial orientation in public organizations . Public Administration 96 ( 4 ): 769 – 86 . OpenURL Placeholder Text WorldCat Slavin , Matt . 2013 . Sustainability in America’s Cities: Creating the Green Metropolis . Washington, DC : Island Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Slothuus , Rune , and Claes H. de Vreese. 2010 . Political parties, motivated reasoning, and issue framing effects . The Journal of Politics 72 ( 3 ): 630 – 45 . OpenURL Placeholder Text WorldCat Swann , William L . 2017 . Modelling the relationship between entrepreneurial orientation, organizational integration, and programme performance in local sustainability . Public Management Review 19 ( 4 ): 542 – 65 . OpenURL Placeholder Text WorldCat Swann , William L. , and Aaron Deslatte. 2019 . What do we know about urban sustainability? A research synthesis and nonparametric assessment . Urban Studies 56 ( 9 ): 1729 – 47 . OpenURL Placeholder Text WorldCat Terman , Jessica N. 2015 . Performance goal achievement in fiscal federalism: The influence of state partisan environments and regulatory regimes . Policy Studies Journal 43 ( 3 ): 333 – 54 . OpenURL Placeholder Text WorldCat Terman , Jessica N. , and Richard C. Feiock. 2015 . Improving outcomes in fiscal federalism: Local political leadership and administrative capacity . Journal of Public Administration Research and Theory 25 ( 4 ): 1059 – 80 . OpenURL Placeholder Text WorldCat Terman , Jessica N. , Anthony Kassekert, Richard C. Feiock, and Kaifeng Yang. 2016 . Walking in the shadow of pressman and Wildavsky: Expanding fiscal federalism and goal congruence theories to single-shot games . The Review of Policy Research 33 ( 2 ): 124 – 39 . OpenURL Placeholder Text WorldCat Teske , Paul , and Mark Schneider. 1994 . The bureaucratic entrepreneur: The case of city managers . Public Administration Review 54 ( 4 ): 331 – 40 . OpenURL Placeholder Text WorldCat Thaler , Richard H. , and L. J. Ganser. 2015 . Misbehaving: The Making of Behavioral Economics. New York : WW Norton . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Tomarken , Andrew J. , and Niels G. Waller. 2005 . Structural equation modeling: Strengths, limitations, and misconceptions . Annual Review of Clinical Psychology 1 : 31 – 65 . OpenURL Placeholder Text WorldCat Tversky , Amos , and Daniel Kahneman. 1981 . The framing of decisions and the psychology of choice . Science 211 ( 4481 ): 453 – 8 . U.S. Government Accountability Office . 2011 . Energy Efficiency and Conservation Block Grant Recipients Face Challenges Meeting Legislative and Program Goals and Requirements , 11 – 379 . Washington, DC : Government Printing Office . https://www.gao.gov/products/GAO-11-379. Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Walker , Richard M. , Frances S. Berry, and Claudia N. Avellaneda. 2015 . Limits on innovativeness in local government: Examining capacity, complexity, and dynamism in organizational task environments . Public Administration 93 ( 3 ): 663 – 83 . OpenURL Placeholder Text WorldCat Wang , Xiaohu , Christopher V. Hawkins, Nick Lebredo, and Evan M. Berman. 2012 . Capacity to sustain sustainability: A study of US cities . Public Administration Review 72 ( 6 ): 841 – 53 . OpenURL Placeholder Text WorldCat Wang , Shu , and Michael A. Pagano. 2017 . Cities and fiscal federalism in the trump era: A discussion . State and Local Government Review 49 ( 3 ): 184 – 98 . OpenURL Placeholder Text WorldCat Workman , Samuel , Bryan D. Jones, and Ashley E. Jochim. 2009 . Information processing and policy dynamics . Policy Studies Journal 37 ( 1 ): 75 – 92 . OpenURL Placeholder Text WorldCat Yi , Hongtao , Rachel M. Krause, and Richard C. Feiock. 2017 . Back-pedaling or continuing quietly? Assessing the impact of ICLEI membership termination on cities’ sustainability actions . Environmental Politics 46 ( 1 ): 138 – 60 . OpenURL Placeholder Text WorldCat Zeemering , Eric S . 2014 . Collaborative Strategies for Sustainable Cities: Economy, Environment and Community in Baltimore . New York : Routledge . © The Author(s) 2020. Published by Oxford University Press on behalf of the Public Management Research Association. All rights reserved. For permissions, please e-mail: journals.permissions@oup.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/open_access/funder_policies/chorus/standard_publication_model)
TI - Performance, Satisfaction, or Loss Aversion? A Meso–Micro Assessment of Local Commitments to Sustainability Programs
JF - Journal of Public Administration Research and Theory
DO - 10.1093/jopart/muaa021
DA - 2020-05-05
UR - https://www.deepdyve.com/lp/oxford-university-press/performance-satisfaction-or-loss-aversion-a-meso-micro-assessment-of-P0dwAeWcpp
SP - 1
VL - Advance Article
IS - 
DP - DeepDyve
ER -