DB.
Tamratyare
ArticleGoalsTrustParticipationandFeedbackLinkingInternalManagementwithPerformanceOutcomes.pdf
© The Author 2014. Published by Oxford University Press on behalf of the Journal of Public Administration Research and Theory, Inc. All rights reserved. For permissions, please e-mail: [email protected].
327
Journal of Public Administration Research And Theory, 2016, 327–343 doi:10.1093/jopart/muu044
Article Advance Access publication October 17, 2014
Article
Goals, Trust, Participation, and Feedback: Linking Internal Management With Performance Outcomes Nathan Favero,* Kenneth J. Meier,*,†, Laurence J. O’Toole Jr.‡,§
*Texas A&M University; †Cardiff University; ‡University of Georgia; §University of Twente
Address correspondence to the author at [email protected].
Abstract
Much recent work in the study of public administration has emphasized new challenges and relatively unusual aspects of management. However, it is likely that the core features of tra- ditional public administration play a crucial role, particularly regarding the delivery of perfor- mance. The most venerable of these aspects of public management have to do with “internal” management. We focus here on a cluster of key, intertwined management practices: setting challenging but feasible goals, building trust through credible commitments, encouraging employee participation, and providing feedback. We examine the relationship between such internal management at the mid-level, as perceived by subordinates rather than the manag- ers themselves, and educational performance for more than 1,100 schools in the New York City school system over a 3-year period. The results indicate that internal management mat- ters, often sizably, for delivering educational outcomes. The findings are robust to autoregres- sive specifications and the purging of halo effects, and they hold across multiple performance measures. Managers’ setting challenging goals appears to be especially important in generat- ing educational results.
Public management matters for program perfor- mance. This basic proposition has been demonstrated in contexts as different as English local governments (Boyne and Walker 2006), Texas school districts (Meier and O’Toole 2003), US state governments (Jacobson, Palus, and Bowling 2010), Colombian local governments (Avellañeda 2009), and local law enforcement agencies (Nicholson-Crotty and O’Toole 2004). Much of the work, however, has focused on how managers deal with the external environment of the organizations (for instance, Agranoff 2007; Provan and Milward 1995). Substantially less work has focused how managers act within an organization
to shape overall program performance. This gap is especially noteworthy given that public managers are frequently restricted in their use of monetary incentives and must rely on normative and solidary inducements to shape employee actions. This study examines how the actions of managers within the organization can generate greater levels of perfor- mance. The empirical evidence is drawn from an elaborate database from New York City schools. The study brings two methodological innovations to the scholarship on management. First, we rely on subor- dinates’ views of what management does rather than managers’ self-reports on their own behavior. Second, we deal with the halo effect of surveys, a measure- ment problem that results because individuals may respond to surveys in a biased manner that is related to their own views of the organization. After a brief
An earlier version of this study was presented at the annual meeting of the American Political Science Association, August 29–September 2, 2012, New Orleans, LA. Copyright American Political Science Association.
328 Journal of Public Administration Research and Theory, 2016, Vol. 26, No. 2
summary of the existing literature, we craft a model of internal management that relies on four concepts— goals, trust, participation, and feedback. This model is then evaluated in the context of public education, and the implications of our findings for the study of public management are discussed.
The Public Management–Performance Nexus
Although the notion that what public managers do can shape the performance of public agencies has long been assumed, large-N empirical research on the man- agement-and-performance question has been largely a product of the past dozen years or so. Early con- tributions were primarily theoretical (Lynn, Heinrich, and Hill 2001; O’Toole and Meier 1999), but empiri- cal analyses followed quickly (see O’Toole and Meier 2011; Rainey 2014 for summaries). Some of these have been based on limited performance measures, such as perceptual assessments by managers themselves1 or federal government–generated PART (Program Assessment Rating Tool) scores, but interesting anal- yses have also been conducted using archival meas- ures of performance, particularly those recorded on a recurring basis and tapping outcomes of salience in the empirical setting. It has been shown that multiple aspects of public management contribute in a variety of ways to public services performance (Boyne et al. 2006; O’Toole and Meier 2011; Rainey 2014).
For all the welcome attention to management’s influence on performance, nonetheless, a great deal of the theoretical and empirical work has dealt with only a portion of public management. Substantial work, for instance, has focused on networks and managerial networking, interorganizational and intergovernmen- tal collaboration, and various other aspects of exter- nally oriented management efforts (Agranoff 2007; Andrews et al. 2010; Jacobson, Palus, and Bowling 2010; Provan and Milward 1995). There has also been much promotion and critique of the so-called New Public Management (Barzelay 1992; Hood 1991; Kettl 2005; O’Toole and Meier 2009). Similarly, additional attention has been devoted to such management topics as the management of information technology, manag- ing for results, or performance management (Ingraham, Joyce, and Donahue 2003; Moynihan 2008).
Relatively neglected in this upsurge of interest in estimating the effects of management on performance have been the tried-and-true functions of internal organizational management, what we might think of as traditional public administration. Of course, the POSDCORB-style elements of traditional public
administration have not been completely ignored.2 However, the effects of many internal management activities—particularly the relationships that manag- ers develop with workers and how such relationships can shape performance—have yet to be explored with systematic data.
Attempting to analyze the performance-related effects of all such aspects of internal management would be a Herculean task. Our approach is to focus on a more limited set of managerial tasks—ones for which there is good reason to believe that there should be performance impacts—and estimate relationships. We focus on four core elements of public management that are interrelated but have received little rigorous empirical attention: managers’ setting and communi- cating goals, their effort to build trust with employees, their enticing worker participation in decision-making processes, and their providing feedback to subordi- nates. These core elements of management envision a people-centered approach to management that relies more on intrinsic motivation than on monetary incen- tives. The aim of this article is to assess what effect— if any—these interconnected aspects of management have on organizational performance. We hypothesize that they positively contribute to that performance. We explore whether this is so in an innovative way by examining the behavior of mid-level managers in organizations devoted to public education, as reported by the managers’ professional subordinates—in this case school teachers.
Goal-Oriented Core Elements of Internal Management
Observers have often commented upon the relatively limited ability possessed by public managers to influ- ence what goes on in and through their organizations (for instance, Wilson 1990). Certainly such manag- ers’ opportunities to control material incentives are much more limited than in the private sector; and in
1 For analyses of the methodological problems associated with a reliance on managers’ perceptions of their own organizations’ performance as a dependent variable, see Meier and O’Toole (2013a, 2013b).
2 POSDCORB stood for Planning, Organizing, Staffing, Directing, Coordinating, Reporting, and Budgeting. The term was popularized by Luther Gulick (1937). Some of these functions have been incorporated into the work conducted under the aegis of the Government Performance Project (Ingraham 2007; Ingraham, Joyce, and Donahue 2003), but that Project did not actually seek to demonstrate a link between management and true organizational performance. Internal public management makes an appearance in the reduced form model of governance proposed by Lynn, Heinrich, and Hill (2001), but managerial effects on performance are not estimated. “Internal management” is also one of the key terms in O’Toole and Meier’s (1999) model of management and performance, although much of their early empirical work focused on the external functions of public managers. Later empirical work made progress on some internal management activities, such as personnel stability, the general management of human capital, and buffering organizations from external perturbations (O’Toole and Meier 2011, 131–81).
329Journal of Public Administration Research and Theory, 2016, Vol. 26, No. 2
the short to medium term, their control over structural features of their organizations is often constrained by law, regulation, and the influence of political princi- pals. Public managers, as a result, need to rely more on normative inducements and on creating a work environment that encourages street-level bureaucrats (see Lipsky 1980; May and Winter 2009; Maynard- Moody and Musheno 2003; Riccucci 2005; Tummers and Bekkers 2014) to adopt the goals of the organiza- tion and enthusiastically pursue these goals. Although a theoretical argument can be made that public man- agers’ influence over their organization’s performance is different in predictable ways from that possessed by private sector managers, we can expect that the influ- ence is nevertheless real (Meier and O’Toole 2011).
Some long-recognized features of internal manage- ment, in particular, would seem to offer possibilities in this regard. At the risk of oversimplifying some of the theoretical details, we specify four elements of man- agement that are common to much of the extant theo- retical work—goals, trust, worker participation, and feedback. We see these four aspects of management as closely interrelated and operating together to enhance organizational performance. Figure 1 summarizes the microfoundations of our theoretical argument linking these four elements of management to organizational performance. Each managerial action seeks to elicit a specific employee behavior. In addition to the primary effect each managerial action exerts on its correspond- ing employee behavior, each managerial element is also expected to exhibit a secondary effect on the other employee behaviors we have identified. This reflects the complementary, interconnected nature of the four sets of managerial actions, which we discuss in greater detail below. Our figure clearly suggests a set of inter- connected relationships too complex to examine in a single study. As such, our empirical test estimates only the effects of the four managerial elements on organi- zational outcomes. Although we do not empirically model the intermediate employee behavior variables, we include them in our figure in order to help clarify the microtheory underlying our hypothesized rela- tionship between internal management actions and performance.
How can managers encourage effective perfor- mance? That is, how does a given managerial action induce employees to engage in behaviors that improve productivity? First, setting clear, feasible yet challeng- ing goals allows employees to understand what man- agement expects of employees and permits employees to focus on efforts that can enhance organizational performance (Latham, Borgogni, and Petitta 2008). Challenging goals are those that require organization members to exert considerable effort to achieve; such goals clearly tap into employees’ intrinsic motivation
to perform well on the job and engage in interest- ing work. Clearly communicated goals should pro- vide direction to employees wishing to improve their knowledge of topics relevant to performing their jobs, and employees can better contribute to policy deci- sions if they have clear goals that help indicate what input might be valued by managers. At the same time, public organizations sometimes have rather ambiguous goals, and ambiguity has been shown to impede per- formance (Chun and Rainey 2006). Even with ambigu- ous goals, managers frequently have the opportunity to communicate with their staff about operational ver- sions of organizational goals, particular objectives to emphasize, and the importance of specific goals for the organization and its stakeholders. Such regularly used communication channels allow for some reduction in ambiguity and initiate a process to establish challeng- ing goals in practice. Goal-setting theory, developed by Locke and coauthors, argues that challenging and spe- cific goals generate better performance than easy and vague ones (for instance, Locke and Latham 1990), and evidence, mostly from the private sector, validates the argument (for instance, Latham 2007; Miner 2005; Pinder 2008).
Second, managers need to engage in actions that build trust among employees. One way to frame this action and the resultant employee behavior is the idea of credible commitment or managerial credibility (see Dull 2009): Managers as principals need to elicit sup- port and cooperation from employees (Barnard 1938), and trust in a manager’s word—belief in her ability and willingness to communicate clearly and follow through on what she says she will do—is central here. This credible commitment can be viewed as establish- ing the legitimacy of the manager and the manager’s actions (Tyler 2006), which is designed to establish trust between management and employees; indeed, trust has been shown to be a crucial element of many institutional settings (Dirks and Ferrin 2002; Ostrom 2005). The need for such commitments is especially the case in terms of managerial reforms because manag- ers are asking for changes in behavior, and employees may be hesitant to change unless they find the promises of the manager credible (Dull 2009, 261). The need for managers to deal with this issue has even been for- mally demonstrated in Gary Miller’s (1993) influential rational choice study of the matter. If trust is estab- lished, employees are clearly more likely to believe what managers say, more likely to pay attention to the priorities that managers articulate and devote time and energy to them, and more likely to both contribute to decision making and be open to receiving feedback from above.
Third, all managers face the problem of information asymmetry. Employees have a great deal of knowledge
330 Journal of Public Administration Research and Theory, 2016, Vol. 26, No. 2
and experience that is valuable in organizational activ- ities. Managers, as a result, need to take actions that elicit participation by workers in decision making in an organization. The potential importance of partici- pation by public employees in organizational deci- sion making has been another venerable theme in the public administration literature (for instance, Mosher 1967) and in some of the most influential studies in the human relations tradition (Argyris 1957). It has also been examined productively in the context of self-determination theory (Gagné and Deci 2005). Participation by employees in highly professionalized organizational settings is likely to be particularly rel- evant, given the likely value of the ideas and creativ- ity of such workers in improving practices and finding ways to enhance outcomes (Huselid 1995; Miller and Monge 1986; Rodgers and Hunter 1991). Such partici- pation can also improve the trust relationship between managers and employees (see Driscoll 1978; Lawler and Hackman 1969). When employees feel that they have a meaningful voice in their organization, they are more likely to be motivated in their work and to buy into organizational goals and learning initiatives.
Fourth, managers need to provide feedback to employees so that employees can receive information about how, and how well, they are contributing to key organizational goals. They can thereby learn to per- form their jobs better and also to focus more on goals that are important to the organization. As Rainey indi- cates, “Difficult goals enhance performance by direct- ing attention and action, mobilizing effort, increasing persistence, and motivating the search for effective performance strategies. Commitment to the goals and
feedback about progress toward achieving them are also necessary for higher performance” (2014, 287, emphasis added; Hrabluik, Latham, and McCarthy 2012; Podsakoff and Farh 1989). These points, which touch upon goals and trust, also clearly indicate the importance of feedback between managers and workers (see also Graber 2003), and so we include this related aspect of management in the present study. Feedback that is both constructive and encouraging can serve not only to inform employees but also to motivate them by providing external validation of their successes and a social incentive to avoid shirking. Delivering helpful and effective feedback may also complement manage- rial efforts to build rapport with employees, increasing employees’ willingness to approach management with their own ideas about how organizational policy or practices might be improved.
In short, a closely related cluster of internally focused managerial actions seek to change the behav- ior of employees so that employee efforts are focused on the goals of the organization, employees trust man- agers and are motivated to work hard, employees are willing to participate in organizational decisions, and employees can learn to perform their jobs better via feedback from management. These core aspects of internal management vary widely across managers and organizations and can be hypothesized to be posi- tively related to organizational performance. Clear goals that are challenging but feasible generate more precise expectations for employees. Building trust among employees via managers’ establishing cred- ibility facilitates a climate encouraging cooperation and the exchange of ideas. The combination of clear
Figure 1. The links between managerial actions, employee behaviors, and organizational performance.
331Journal of Public Administration Research and Theory, 2016, Vol. 26, No. 2
goals and trust encourages worker participation and enhances employees’ willingness to receive construc- tive feedback. All these actions should contribute to more focused employees and a more productive work environment. This interrelationship of the various managerial actions and employee behaviors is fore- shadowed in the work of Barnard (1938, 86) who notes that cooperation by employees and the accept- ance of a goal are essentially a simultaneous activity; the same can be said for trust, participation, and feed- back. In the field of public education, the setting we examine empirically in this investigation, the role of mid-level managers—school principals—is likely to be especially crucial.
The Research Setting
This study examines the performance of New York City schools between 2007 and 2009. New York City has the largest school system in the country with 1.1 million students, 80,000 teachers, and a $21 bil- lion budget. During this time period, the New York City schools were undergoing a major reform led by Superintendent Joel Klein, who was being advised by management scholar William Ouchi (2009) among others. The reforms greatly expanded the author- ity of principals and significantly increased schools’ accountability for performance. In addition to the generic reforms of high standards and accountability, the New York City reforms stressed parental involve- ment (including the establishment of annual citywide parent and student surveys) and the engagement of teachers. The latter is directly relevant to this study given our use of teachers’ assessments of manage- ment. Although aspects of this reform are unique to New York City, virtually all urban school districts are undergoing major reform efforts, often repeatedly (Hess 1999).
Performance indicators and several measures of basic school traits are publicly available from city and state records (New York City Department of Education 2011b; New York State Education Department 2011; New York State Testing and Accountability Reporting Tool 2011). These records were merged with school-level results from an annual survey of teachers conducted each spring by the New York City Department of Education (2011b). All teachers in the school system were given the oppor- tunity to respond to the survey, and the average response rate for the schools we examine was 63%. For the questions we analyze, the city aggregated the responses by calculating the average response in each school (strongly agree = 10, agree = 6.6, disa- gree = 3.3, and strongly disagree = 0). We consider data from 1,164 schools over these 3 years, yielding
a total of 3,267 observations after dropping observa- tions with missing values.3
The primary dependent variable is student perfor- mance on state standardized tests. For elementary and middle schools, we created an index based on 3rd–8th grade English and math scores. We first calculated the average scores as well as the proportion of scores that met proficiency goals for English and math. These four variables were standardized and then summed to form the elementary/middle school index. High schools do not administer annual English and math exams, but they require students to pass several exams, including an English exam and at least one math exam, in order to receive a diploma (New York City Department of Education 2011a). Using state records, we divided the total number of English and math scores that met Regents diploma standards by the schools’ total high school enrollment. We then combined English and math into a single category by adding standardized versions of the two ratios we just produced. This gave us our high school performance index. We combined standardized versions of the elementary/middle school and high school performance indexes into a single variable.4
Several control variables aid us in isolating the unique effect of internal management on school per- formance. Student demographic characteristics indi- cate task difficulty and are important predictors of performance. City records provide data on the percent- age of students who hold various racial/ethnic identi- ties, are female, are recent immigrants, and are special education students. The percentage of Limited English Proficiency students can be found in state records. We measure socioeconomic status with four indicators that allow us to approximate different levels of eco- nomic hardship; the city supplies the percentage of stu- dents in temporary housing and below the poverty line, whereas state records offer the percentage of students
3 Our data set did not include special education schools, alternative schools, charter schools, early childhood schools, transfer schools, or Young Adult Borough Centers. Additionally, we excluded observations where fewer than three teachers or fewer than five parents responded to the survey. We also omitted observations when the variables derived from government records contained missing values or obvious data errors, such as percentages greater than 100. New York City records did not always distinguish between missing values and values of zero; in such cases, we assumed a value of zero was appropriate. We faced a choice between accepting any biases generated as a result of missing data versus seeking to engage in multiple imputation of the data for these cases. Because most of the omitted cases are so different from the average school (and do not really belong in our population), we felt that the distortion created by including them via imputation would be more problematic than the bias generated by omitting them.
4 Because multiple math exams were offered in some years, we had to sum the number of passing scores from each individual math exam to find the total number of passing math scores.
332 Journal of Public Administration Research and Theory, 2016, Vol. 26, No. 2
eligible for free lunch and reduced-price lunch. We also include dummy variables indicating whether or not the school serves elementary, middle school, and high school students (the categories are not mutually exclusive).
Research has shown that class size (Nye, Hedges, and Konstantopoulos 2000) and teacher turnover (Meier and Hicklin 2008) are also important predic- tors of student performance. A single measurement of class size was created that combines several state records of average class size for different grade levels and subject areas.5 We replicated a measure of teacher turnover used by Favero and Meier (2013) for these data. The teacher turnover variable is one of three factor scores produced by a rotated factor analysis of several school-level measures of teacher characteristics (Supplementary Table A2). The turnover factor reflects both the turnover rate of all teachers and the turnover rate among teachers with fewer than 5 years experi- ence. Finally, our models include fixed effects for years and standard errors clustered by school to deal with serial correlation and heteroscedasticity.
Measuring Management
As explained earlier, internal management of pub- lic organizations focuses heavily, of necessity, on the “human side” (McGregor 1960), simply because a variety of structural and incentive-focused elements in management are limited in the public sector (see Feeney and Rainey 2010). One problem with measuring these aspects of management via surveys of managers is that managers might bias their responses. These individu- als are trained managers and have taken courses in the area; they are quite likely to know the literature well. As a result, their responses might be subject to social desirability bias as they seek to demonstrate to the researcher that they are adopting the best current practices in management. Even without social desir- ability bias, what managers think they are doing and what their subordinates think they are doing might be two different things. The manager might not be effec- tive in communicating his or her behavior or might not be as consistent as the survey might indicate. For these reasons, measuring management actions by using the
perceptions of the individuals who are being managed has the potential of generating more unbiased meas- ures of management. All measures of management, as a result, rely on teachers’ perceptions of how the school is managed. The use of teachers’ perceptions of man- agement has another advantage in that it incorporates the time lag that is likely to occur between managerial action and employee response. These perceptions of managerial action capture when an action has begun to have an impact on frontline employees (e.g., the level of trust increases) rather than when the manager initially took the action.
All organizations have goals, and clearly communi- cating goals is important if managers wish for employ- ees to adopt these goals in a meaningful way. The content of these goals also matters, as explained earlier. In the substantive case being examined here, K-12 edu- cation, an extensive literature finds that student per- formance is enhanced by setting high expectations for all students, providing clear measures of performance, and focusing priorities on having students meet these high expectations; high expectations, in fact, might even be the overarching goal under which other goals of education are subsumed (Costrell 1994; Darling- Hammond and Ball 1998; Deci et al. 1991; McCombs and Whisler 1997; Mirel and Angus 1994; Owens and Valesky 2011). A goals measure is created by factor analyzing four items related to goals and high expecta- tions; the survey items include adjectives like “high” and “challenging,” thus linking well to the theoretical argument about challenging goals. This measure has an eigenvalue of 3.66, accounting for 91% of the total variance in these items (table 1). This measure does not explicitly tap goal feasibility, but it otherwise fits the theoretical argument well.
Trust is a concern in all principal-agent relation- ships. How do agents know that the principal will honor the commitments made to them, given that organizations are hierarchies and thus reflect inequi- ties of power (Miller 1993)? The link between cred- ible commitment by the manager and employee trust in organizations has three parts: the belief that the manager represents the interests of the organization, rather than his or her own personal interests; clear communication by the manager; and the develop- ment of trust in the manager by subordinates. Trust/ commitment in our analysis is measured by a factor analysis of seven items (table 1)—two that relate to representing the interests of the organization (aligned curriculum and the needs of students), three that focus on clarity of communication by management, and two that deal with trust and effectiveness. The seven items load on a single factor with an eigenvalue of 6.10, accounting for 87% of the total variance in these items.
5 The state of New York provides a total of nine measurements of average class size: one for all elementary school classes and separate measurements of math, English, science, and social studies classes for both 8th and 10th grade. We converted the state measurements into z-scores within the nine categories. We then used the standardized values to calculate the 8th grade average and the 10th grade average across the four subject areas. Using these two averages along with the standardized elementary school class size, we calculated a final weighted average. The weights for the three variables were based on the share of students enrolled in elementary, middle, and high school grade levels at each school, as documented in state records.
333Journal of Public Administration Research and Theory, 2016, Vol. 26, No. 2
Modern organizations rely on the enthusiastic coop- eration of their members, a basic idea that is recog- nized by virtually all theories of management, ranging from the classic exposition of Barnard (1938) to the decision-theoretic work of Simon (1997) to the semi- nal studies in human relations (for instance, Argyris 1957). By encouraging participation by street-level bureaucrats (Lipsky 1980) in the management process, the manager benefits from the greater range of exper- tise and experience in the organization. Workers who feel their ideas and participation are valued are likely to make a greater commitment to the organization. Participation is likely to be especially important in decentralized, professionalized organizations such as public schools. Three items tap, respectively, whether teachers perceive that the principal has confidence in the expertise of teachers, whether teachers see them- selves as playing a meaningful role in setting goals and making decisions, and whether teachers believe that
school leaders encourage collaboration among teach- ers (table 1). A single factor with an eigenvalue of 2.69 accounts for 90% of the total variance in these items.
Feedback from managers to workers is an essential element of management. Feedback permits goal clarifi- cation, particularly in organizations that have multiple or imprecise goals. Three survey questions probe vari- ous aspects of feedback—classroom visits by manag- ers, feedback about teaching, and the perception that managers place a high priority on quality teaching (table 1). The three items all load on a single factor with an eigenvalue of 2.67, accounting for 89% of the total variance.
Each of the four factors has strong internal reli- ability, as demonstrated by the large amount of com- mon variance accounted for by the factors. Each has been presented as a separate dimension, and in theory, management might sometimes act on one or two of these dimensions but not all four. At the same time,
Table 1. Factor Analyses for Management Measures
Survey Item Factor Loading
Goals My school has high expectations for all students 0.94 My school has clear measures of progress for student achievement throughout the year 0.95 This school makes it a priority to help students develop challenging learning goals 0.97 This school makes it a priority to help students find the best ways to achieve their learning goals 0.96 Eigenvalue 3.66 Proportion 0.91 N 4,341 Trust/commitment School leaders communicate a clear vision for this school 0.96 School leaders let staff know what is expected of them 0.95 School leaders encourage open communication on important school issues 0.94 Curriculum, instruction, and assessment are aligned within and across the grade levels at this school 0.84 The principal places the learning needs of children ahead of other interests 0.95 The principal is an effective manager who makes the school run smoothly 0.96 I trust the principal at his/her word 0.94 Eigenvalue 6.10 Proportion 0.87 N 4,342 Participation The principal has confidence in the expertise of the teachers 0.94 School leaders invite teachers to play a meaningful role in setting goals and making important
decisions for this school 0.96
School leaders encourage collaboration among teachers 0.94 Eigenvalue 2.69 Proportion 0.90 N 4,339 Feedback School leaders visit classrooms to observe the quality of teaching at this school 0.94 School leaders give me regular and helpful feedback about my teaching 0.95 School leaders place a high priority on the quality of teaching at this school 0.94 Eigenvalue 2.67 Proportion 0.89 N 4,339
334 Journal of Public Administration Research and Theory, 2016, Vol. 26, No. 2
an internal management–focused theory would sug- gest that all of these intertwined factors need to be present to effectively manage employees. In support of this integrated approach, we should note that the four factors are highly correlated (the average correlation is .85 with a low of .78 between goals and participa- tion and a high of .92 between trust/commitment and participation). To determine whether the individual factors might operate alone or if they have to be con- sidered in concert, we create a separate overall meas- ure of internal management by factor analyzing all 17 items. Again this factor analysis (Supplementary Table A1) produces a single factor with an eigenvalue of 13.63, accounting for 80% of the variance in all the items. Although this is an impressive set of load- ings, the overall variance explained is less than what the four individual items explain. This result suggests that examining both the individual items and the over- all item is a sensible approach. This analysis needs to proceed with some caution, since the high collinear- ity among the individual items can lead to misleading results. To verify the sets of results, the individual items will be used by themselves one at a time, in a single equation with all four, and also in an equation with the combined overall measure.
Findings
The second model in table 2 shows that the combined measure of internal management is strongly and posi- tively related to student performance, as measured by test scores. Both variables are in standardized units, so a one standard deviation (SD) change in internal man- agement is associated with 0.122 SD change in student test scores. Although this is not a massive change, over the full range of the data, the total impact could be as large as 0.73 SDs, a change that would be substan- tively very important. The assessment of the individual items illustrates the problem of collinearity among the internal management variables. Supplementary Table A3 shows that when each of the four variables (goals, trust/commitment, participation, and feedback) is included in the equation by itself, each is positively and significantly related to student performance. When they are all included in the same equation, as in Model 1 (table 2), the measures get no credit for their com- mon variation, only their unique variation. This analy- sis is revealing, in that the impact of goals appears to have value over and above the other aspects of internal management by itself, while trust’s unique aspect is negative and significant.
The control variables in our initial models in table 2 all behave as predicted, with the exception of the posi- tive relationship for the percentage of students qualify- ing for reduced-price school lunches. That relationship
is an interesting anomaly. It is not the result of collin- earity but simply an indication that schools with many students in households with this modest income indi- cator actually outperform schools with students who are not eligible for reduced-price school lunches and those with more students on free lunch. Teacher turno- ver is set off from the other controls simply because management is likely to affect turnover (see below). It also provides an interesting comparison to the effect size of the management variable. When adjusted for its SD, the impact of teacher turnover is approximately half the size of the impact of management. This rela- tive comparison provides another substantive inter- pretation of the effect size. Because our concern with including the control variables was to ensure that the model was well specified (our first two models explain between 71% and 73% of the variance in student per- formance), we will not discuss the individual relation- ships for these variables other than teacher turnover.
Autoregressive Models To provide an even more rigorous test of the impact of internal management on performance, Models 3 and 4 repeat these analyses while also controlling for the school’s performance in the previous year. Schools, like all bureaucracies, are highly autoregressive systems. Their processes and personnel change only slowly, and most of their students continue in the same school, year after year. As one might expect, the level of explained variance jumps dramatically when the lagged depend- ent variable is added to the model. Further, these estimations provide a tougher test for the impact of management on performance, since they control for past (last year’s) performance. The combined meas- ure of management remains positive and statistically significant. Although the size of the coefficient drops, this coefficient must now be interpreted as the impact of management on student performance this year over and above any impact last year; in brief, this is the short-term impact. At the same time, improvements in performance one year as the result of management then become part of the base level of performance the next year, so that the impact of management continues on into the future in what is known as a geometrically distributed lag (Pindyck and Rubinfeld 1991, 204–8). The long-term impact of a one unit change in manage- ment can be calculated by using the coefficient along with the coefficient for the lagged dependent variable.6 When this is done, the long-term impact is estimated to be 0.16 or actually larger than the estimate in Model 2.
6 The long-run impact can be estimated by dividing the regression coefficient by one minus the coefficient for the lagged dependent variable.
335Journal of Public Administration Research and Theory, 2016, Vol. 26, No. 2
Ta b
le 2
. M
o d
e ls
o f
S tu
d e n
t P
e rf
o rm
a n
ce
M o d el
1 M
o d el
2 M
o d
el 3
M o
d el
4
b (
SE )
b (
SE )
b (
SE )
b (
SE )
L ag
ge d s
tu d
en t
p er
fo rm
an ce
0 .8
1 1
* *
* (
0 .0
0 9
) 0
.8 1
8 *
* *
( 0
.0 0
9 )
In te
rn al
m an
ag em
en t
0 .1
2 2 * * * (
0 .0
1 2 )
0 .0
2 9
* *
* (
0 .0
0 4
)
G o al
s 0 .2
6 3 * * * (
0 .0
2 9 )
0 .0
6 1
* *
* (
0 .0
1 0
)
T ru
st /c
o m
m it
m en
t − 0 .1
3 8 * * * (
0 .0
3 6 )
− 0
.0 2
2 (
0 .0
1 2
)
P ar
ti ci
p at
io n
0 .0
2 1 (
0 .0
2 8 )
− 0
.0 0
6 (
0 .0
0 9
)
F ee
d b ac
k 0 .0
2 8 (
0 .0
2 7 )
0 .0
0 9
( 0
.0 0
9 )
T ea
ch er
t u rn
o ve
r − 0 .0
5 7 * * * (
0 .0
1 0 )
− 0 .0
5 1 * * * (
0 .0
1 0 )
− 0
.0 1
5 *
* *
( 0
.0 0
4 )
− 0
.0 1
3 *
* (
0 .0
0 4
) A
m er
ic an
/A la
sk an
n at
iv e
− 0 .0
6 9 * * * (
0 .0
2 1 )
− 0 .0
6 8 * * (
0 .0
2 1 )
− 0
.0 1
0 (
0 .0
0 8
) −
0 .0
1 0
( 0
.0 0
8 )
B la
ck − 0 .0
0 8 * * * (
0 .0
0 1 )
− 0 .0
0 9 * * * (
0 .0
0 1 )
− 0
.0 0
1 *
( 0
.0 0
0 )
− 0
.0 0
1 *
* (
0 .0
0 0
) H
is p an
ic − 0 .0
0 5 * * * (
0 .0
0 1 )
− 0 .0
0 6 * * * (
0 .0
0 1 )
0 .0
0 0
( 0
.0 0
0 )
0 .0
0 0
( 0
.0 0
0 )
A si
an 0 .0
0 4 * * * (
0 .0
0 1 )
0 .0
0 4 * * * (
0 .0
0 1 )
0 .0
0 1
* (
0 .0
0 0
) 0
.0 0
1 *
( 0
.0 0
0 )
F em
al e
0 .0
0 3 (
0 .0
0 2 )
0 .0
0 4 * (
0 .0
0 2 )
0 .0
0 1
( 0
.0 0
0 )
0 .0
0 1
( 0
.0 0
0 )
L im
it ed
E n gl
is h
− 0 .0
1 1 * * * (
0 .0
0 2 )
− 0 .0
1 1 * * * (
0 .0
0 2 )
− 0
.0 0
2 *
* *
( 0
.0 0
1 )
− 0
.0 0
2 *
* *
( 0
.0 0
1 )
R ec
en t
im m
ig ra
n ts
− 0 .0
0 7 (
0 .0
0 6 )
− 0 .0
0 8 (
0 .0
0 7 )
0 .0
0 1
( 0
.0 0
2 )
0 .0
0 1
( 0
.0 0
2 )
Sp ec
ia l ed
u ca
ti o n
− 0 .0
1 8 * * * (
0 .0
0 2 )
− 0 .0
2 1 * * * (
0 .0
0 3 )
− 0
.0 0
1 (
0 .0
0 1
) −
0 .0
0 1
( 0
.0 0
1 )
T em
p o ra
ry h
o u si
n g
− 0 .0
0 4 (
0 .0
0 2 )
− 0 .0
0 4 (
0 .0
0 3 )
0 .0
0 3
* (
0 .0
0 1
) 0
.0 0
3 *
( 0
.0 0
1 )
P o ve
rt y
ra te
− 0 .0
0 5 * * * (
0 .0
0 1 )
− 0 .0
0 5 * * * (
0 .0
0 1 )
− 0
.0 0
0 (
0 .0
0 0
) −
0 .0
0 0
( 0
.0 0
0 )
F re
e lu
n ch
− 0 .0
0 1 (
0 .0
0 1 )
− 0 .0
0 1 (
0 .0
0 1 )
− 0
.0 0
0 (
0 .0
0 0
) −
0 .0
0 0
( 0
.0 0
0 )
R ed
u ce
d l u n ch
0 .0
0 6 * (
0 .0
0 2 )
0 .0
0 7 * * (
0 .0
0 2 )
− 0
.0 0
0 (
0 .0
0 1
) −
0 .0
0 0
( 0
.0 0
1 )
Y ea
r 2 0 0 8
0 .1
7 9 * * * (
0 .0
1 1 )
0 .2
2 2 * * * (
0 .0
1 0 )
0 .0
9 1
* *
* (
0 .0
1 0
) 0
.1 0
1 *
* *
( 0
.0 0
9 )
Y ea
r 2 0 0 9
0 .4
8 7 * * * (
0 .0
1 8 )
0 .5
5 6 * * * (
0 .0
1 8 )
0 .1
8 5
* *
* (
0 .0
1 1
) 0
.1 9
8 *
* *
( 0
.0 1
0 )
E le
m en
ta ry
s ch
o o l
0 .4
1 0 * * * (
0 .0
4 7 )
0 .4
6 9 * * * (
0 .0
4 7 )
0 .0
1 6
( 0
.0 1
5 )
0 .0
2 5
( 0
.0 1
4 )
M id
d le
s ch
o o l
− 0 .1
6 3 * * * (
0 .0
3 0 )
− 0 .1
8 7 * * * (
0 .0
3 1 )
0 .0
4 6
* *
* (
0 .0
1 1
) 0
.0 4
1 *
* *
( 0
.0 1
1 )
H ig
h s
ch o o
l 0 .0
5 9 (
0 .0
4 9 )
0 .0
0 6 (
0 .0
5 0 )
− 0
.2 1
3 *
* *
( 0
.0 1
5 )
− 0
.2 2
9 *
* *
( 0
.0 1
5 )
A ve
ra ge
c la
ss s
iz e
− 0 .0
0 2 (
0 .0
1 2 )
− 0 .0
0 8 (
0 .0
1 2 )
− 0
.0 1
0 *
( 0
.0 0
4 )
− 0
.0 1
1 *
* (
0 .0
0 4
) C
o n st
an t
0 .7
8 2 * * * (
0 .1
3 7 )
0 .8
3 4 * * * (
0 .1
3 7 )
0 .2
0 5
* *
* (
0 .0
3 5
) 0
.2 1
0 *
* *
( 0
.0 3
4 )
A d ju
st ed
R 2
0 .7
2 5
0 .7
1 3
0 .9
4 4
0 .9
4 3
N 3 ,2
6 7
3 ,2
6 7
3 ,2
6 7
3 ,2
6 7
N o te
: * p <
. 0 5 ; * * p <
. 0 1 ; * * * p <
. 0 0 1 .
336 Journal of Public Administration Research and Theory, 2016, Vol. 26, No. 2
The collinearity of the four individual management items evident before, of course, will affect the results in the autoregressive equations. Each of the individual items when entered by itself is positively and signifi- cantly related to student performance (Supplementary Table A4). This individual impact is clearly the result of the shared variance, as Model 3 shows that only the goal measure remains significant when all four are in the model, again suggesting that the stress on high expectations and goals contributes some unique ele- ments to internal management.
Addressing the Halo Effect We can go further in exploring the management-and- performance relationship by dealing with a meas- urement issue. Although employee perceptions of management are preferable to self-reported (i.e., man- ager-reported) measures in many ways, they are not perfect (there are no perfect measures). Employee per- ceptions can suffer from halo effects. In their simplest form, halo effects refer to the tendency of respond- ents’ overall perceptions to influence assessments of specific traits (Cooper 1981; Thorndike 1920). In our case, employee assessments of management might be colored by their broader opinions about the organiza- tion and its performance.
To determine whether the findings in this study are the result of general biases generated by halo effects, we can control for such halo effects by, first, factor analyzing all survey questions answered by the teach- ers that relate to performance. The first common fac- tor will yield a good indication of the teacher’s overall perception of the school, including any halo effect (any positivity or negativity bias). It will also contain some real variation related to management, performance, and other important factors (Supplementary Table A5). Next, we purge the key management variables of the halo effect by regressing each one on the common factor and taking the residuals from these regressions as the new measures of management (Supplementary Table A6). We then substitute these halo-corrected measures of management into the regression equa- tions we ran before. This test overcorrects and thus risks throwing away real variation in an attempt to eliminate the halo effect. Whereas the models with uncorrected measures may have overstated the effect of management on performance, the halo-corrected measures should understate this same effect. Therefore, the halo-corrected models make obtaining significant findings difficult; but whatever findings remain are unlikely to be spurious owing to the measurement bias of any halo effects.
Table 3 presents the halo-corrected results for management. When this adjustment is done for the overall measure, the findings are very similar to what
they were in the previous analysis. Internal manage- ment is positively associated with performance both with and without including a lagged dependent vari- able. The halo correction does reduce the size of the coefficient so that its effect size is now approximately equal to the impact of teacher turnover. The clear conclusion for the overall measure, nonetheless, is that the results are not spurious owing to any halo effect from the survey. The individual items, when used simultaneously, generate similar results without the lagged dependent variable, but with the lagged dependent variable, the goals measure is now only significant when using a one-tailed test (the t-score is less than 1.96 but more than 1.65). When the four management measures are entered individually into the equation (Supplementary Tables A7 and A8), all relationships remain positive and statistically significant.7
The Interactive Effects of Management and Goals The pattern of relationships for the overall (internal management) measure and the goals measure suggests that an effort be made to deal with the high collinear- ity of the two measures and to assess the relationships together. After all, one might expect that the combi- nation of high standards for students and effective management more broadly might mean more in com- bination than either the high standards for students or effective management would produce separately. To avoid the problems of collinearity, we regress the goals index on the overall internal management index and take the residuals. These residuals capture the part of the goals measure that is independent of the over- all internal management measure. Model 9 in table 4 shows that both measures are strongly positive in pre- dicting performance; in fact, when one adjusts these coefficients for the differences in the variance of the two measures, they both have approximately the same effect size (0.13). Model 10 includes a term represent- ing the interaction of goals and the remaining aspects of internal management, and it confirms the hypothesis that high expectations (goals) and good management have greater impact when both are present. To illus- trate, at 2 SDs above the mean for goals, the impact of internal management doubles; at 2 SDs below the mean for goals, the impact of internal management is essentially zero. Internal management matters more when goals are challenging but contributes little to the attainment of easy goals. This finding is consistent with
7 As a final robustness check, we estimated the models using two-way fixed effects to control for any possible omitted variable bias. Although these models at times reduced the relative size of the coefficients, the models produced statistically significant results for both internal management and goals for all six performance indicators used in this analysis.
337Journal of Public Administration Research and Theory, 2016, Vol. 26, No. 2
Ta b
le 3
. M
o d
e ls
o f
S tu
d e n
t P
e rf
o rm
a n
ce (
w it
h h
a lo
c o
rr e ct
io n
s)
M o d el
5 M
o d el
6 M
o d
el 7
M o
d el
8
b (
SE )
b (
SE )
b (
SE )
b (
SE )
L ag
ge d s
tu d
en t
p er
fo rm
an ce
0 .8
2 5
* *
* (
0 .0
0 9
) 0
.8 2
6 *
* *
( 0
.0 0
9 )
In te
rn al
m an
ag em
en t
(h al
o c
o rr
ec ti
o n s)
0 .0
7 5 * * * (
0 .0
1 3 )
0 .0
1 5
* *
* (
0 .0
0 4
)
G o al
s (h
al o c
o rr
ec ti
o n s)
0 .1
3 2 * * * (
0 .0
3 0 )
0 .0
2 0
( 0
.0 1
0 )
T
ru st
/c o m
m it
m en
t (h
al o c
o rr
ec ti
o n s)
− 0 .1
0 6 * * (
0 .0
3 7 )
− 0
.0 1
1 (
0 .0
1 2
)
P ar
ti ci
p at
io n (
h al
o c
o rr
ec ti
o n s)
0 .0
4 2 (
0 .0
2 9 )
− 0
.0 0
2 (
0 .0
0 9
)
F ee
d b ac
k (
h al
o c
o rr
ec ti
o n s)
0 .0
4 2 (
0 .0
2 8 )
0 .0
1 3
( 0
.0 0
9 )
T ea
ch er
t u rn
o ve
r − 0 .0
6 0 * * * (
0 .0
1 0 )
− 0 .0
5 7 * * * (
0 .0
1 0 )
− 0
.0 1
5 *
* *
( 0
.0 0
4 )
− 0
.0 1
5 *
* (
0 .0
0 4
) A
m er
ic an
/A la
sk an
n at
iv e
− 0 .0
7 1 * * * (
0 .0
2 1 )
− 0 .0
6 9 * * (
0 .0
2 1 )
− 0
.0 1
0 (
0 .0
0 8
) −
0 .0
1 0
( 0
.0 0
8 )
B la
ck − 0 .0
1 0 * * * (
0 .0
0 1 )
− 0 .0
1 0 * * * (
0 .0
0 1 )
− 0
.0 0
1 *
* (
0 .0
0 0
) −
0 .0
0 1
* *
( 0
.0 0
0 )
H is
p an
ic − 0 .0
0 5 * * * (
0 .0
0 1 )
− 0 .0
0 6 * * * (
0 .0
0 1 )
0 .0
0 0
( 0
.0 0
0 )
0 .0
0 0
( 0
.0 0
0 )
A si
an 0 .0
0 3 * * (
0 .0
0 1 )
0 .0
0 4 * * (
0 .0
0 1 )
0 .0
0 0
( 0
.0 0
0 )
0 .0
0 0
( 0
.0 0
0 )
F em
al e
0 .0
0 5 * (
0 .0
0 2 )
0 .0
0 5 * (
0 .0
0 2 )
0 .0
0 1
* (
0 .0
0 0
) 0
.0 0
1 *
( 0
.0 0
0 )
L im
it ed
E n gl
is h
− 0 .0
1 1 * * * (
0 .0
0 2 )
− 0 .0
1 2 * * * (
0 .0
0 2 )
− 0
.0 0
2 *
* *
( 0
.0 0
1 )
− 0
.0 0
2 *
* *
( 0
.0 0
1 )
R ec
en t
im m
ig ra
n ts
− 0 .0
0 7 (
0 .0
0 7 )
− 0 .0
0 7 (
0 .0
0 7 )
0 .0
0 1
( 0
.0 0
2 )
0 .0
0 2
( 0
.0 0
2 )
Sp ec
ia l ed
u ca
ti o n
− 0 .0
2 1 * * * (
0 .0
0 3 )
− 0 .0
2 2 * * * (
0 .0
0 3 )
− 0
.0 0
1 (
0 .0
0 1
) −
0 .0
0 1
( 0
.0 0
1 )
T em
p o ra
ry h
o u si
n g
− 0 .0
0 4 (
0 .0
0 3 )
− 0 .0
0 4 (
0 .0
0 3 )
0 .0
0 3
* (
0 .0
0 1
) 0
.0 0
3 *
( 0
.0 0
1 )
P o ve
rt y
ra te
− 0 .0
0 5 * * * (
0 .0
0 1 )
− 0 .0
0 5 * * * (
0 .0
0 1 )
− 0
.0 0
0 (
0 .0
0 0
) −
0 .0
0 0
( 0
.0 0
0 )
F re
e lu
n ch
− 0 .0
0 1 (
0 .0
0 1 )
− 0 .0
0 1 (
0 .0
0 1 )
− 0
.0 0
0 (
0 .0
0 0
) −
0 .0
0 0
( 0
.0 0
0 )
R ed
u ce
d l u n ch
0 .0
0 7 * * (
0 .0
0 3 )
0 .0
0 7 * * (
0 .0
0 3 )
− 0
.0 0
0 (
0 .0
0 1
) −
0 .0
0 0
( 0
.0 0
1 )
Y ea
r 2 0 0 8
0 .2
7 3 * * * (
0 .0
0 9 )
0 .2
7 4 * * * (
0 .0
0 9 )
0 .1
1 2
* *
* (
0 .0
0 9
) 0
.1 1
2 *
* *
( 0
.0 0
9 )
Y ea
r 2 0 0 9
0 .6
1 8 * * * (
0 .0
1 7 )
0 .6
2 6 * * * (
0 .0
1 7 )
0 .2
1 0
* *
* (
0 .0
1 1
) 0
.2 1
1 *
* *
( 0
.0 1
1 )
E le
m en
ta ry
s ch
o o l
0 .4
6 9 * * * (
0 .0
4 8 )
0 .4
8 6 * * * (
0 .0
4 9 )
0 .0
2 2
( 0
.0 1
5 )
0 .0
2 4
( 0
.0 1
5 )
M id
d le
s ch
o o l
− 0 .1
9 5 * * * (
0 .0
3 2 )
− 0 .2
0 2 * * * (
0 .0
3 2 )
0 .0
4 2
* *
* (
0 .0
1 1
) 0
.0 4
0 *
* *
( 0
.0 1
1 )
H ig
h s
ch o o
l − 0 .0
2 1 (
0 .0
5 1 )
− 0 .0
2 9 (
0 .0
5 1 )
− 0
.2 3
7 *
* *
( 0
.0 1
5 )
− 0
.2 3
9 *
* *
( 0
.0 1
5 )
A ve
ra ge
c la
ss s
iz e
− 0 .0
1 4 (
0 .0
1 2 )
− 0 .0
1 4 (
0 .0
1 2 )
− 0
.0 1
3 *
* (
0 .0
0 4
) −
0 .0
1 3
* *
( 0
.0 0
4 )
C o n st
an t
0 .7
5 7 * * * (
0 .1
4 0 )
0 .8
1 1 * * * (
0 .1
4 0 )
0 .1
9 5
* *
* (
0 .0
3 4
) 0
.2 0
1 *
* *
( 0
.0 3
4 )
A d ju
st ed
R 2
0 .7
0 5
0 .7
0 1
0 .9
4 2
0 .9
4 2
N 3 ,2
6 7
3 ,2
6 7
3 ,2
6 7
3 ,2
6 7
N o te
: * p <
. 0 5 ; * * p <
. 0 1 ; * * * p <
. 0 0 1 .
338 Journal of Public Administration Research and Theory, 2016, Vol. 26, No. 2
the idea that organizations should assign their best managers to the most challenging problems.
Alternative Measures of Performance Organizations have multiple goals and multiple ways of assessing performance, and schools are no excep- tion to this rule. The stakeholders of public school systems typically care about more than standardized examination scores, and so do those who manage the schools. As a final step in this analysis, we explore the relationship between internal management and some additional objectives. To investigate the impact of internal management on various other goals that the schools can be expected to pursue, we include five more output and outcome measures—parental sat- isfaction, student attendance, the official education progress report, a measure of school violence, and teacher turnover. Interestingly, these goals, along with schools’ performance on the standardized tests, are not as highly correlated with one another as some might expect. The largest correlation (−.49) is between the measure of school violence and the test score measure, with numerous bivariate correlations between goals considerably lower than that. At most, therefore, these goal pairs share less than 25% common variance. We
examine whether management contributes to these additional goals while using the same set of controls for these equations as we did for the earlier analyses.
Parental Satisfaction New York City schools annually survey parents in the spring to ask them about their evaluation of the public schools that their children attend. These annual surveys have a response rate of 43% and ask a wide variety of performance questions. Using six survey indicators, we created a factor score that rep- resented how favorably parents evaluated each school (Supplementary Table A9). The factor had an eigen- value of 4.96 and accounted for 83% of the varia- tion in the individual items. Table 5 shows that the measure of internal management is positively associ- ated with parental satisfaction. Because this measure is also a factor score, the relative impact of manage- ment on parent satisfaction appears to be larger than the impact on student performance. The goals meas- ure (the residuals-based measure) has a similar strong and positive impact on parental evaluations of the schools, and this effect is again much larger for paren- tal assessments than for the test score measure. The interaction of the management and the goals variable
Table 4. Models of Student Performance (with interactive term)
Model 9 Model 10
b (SE) b (SE)
Management 0.150*** (0.012) 0.148*** (0.012) Goals (residual measure) 0.296*** (0.036) 0.340*** (0.040) Management × goals (residual measure) 0.076* (0.030) Teacher turnover −0.057*** (0.010) −0.059*** (0.009) American/Alaskan native −0.068*** (0.021) −0.067** (0.021) Black −0.008*** (0.001) −0.008*** (0.001) Hispanic −0.005*** (0.001) −0.005*** (0.001) Asian 0.004*** (0.001) 0.004*** (0.001) Female 0.003 (0.002) 0.003 (0.002) Limited English −0.011*** (0.002) −0.011*** (0.002) Recent immigrants −0.007 (0.006) −0.007 (0.006) Special education −0.018*** (0.002) −0.017*** (0.002) Temporary housing −0.004 (0.003) −0.005 (0.002) Poverty rate −0.005*** (0.001) −0.005*** (0.001) Free lunch −0.001 (0.001) −0.001 (0.001) Reduced lunch 0.006* (0.002) 0.006* (0.002) Year 2008 0.175*** (0.011) 0.174*** (0.011) Year 2009 0.485*** (0.018) 0.483*** (0.018) Elementary school 0.414*** (0.047) 0.408*** (0.047) Middle school −0.166*** (0.030) −0.165*** (0.030) High school 0.065 (0.048) 0.072 (0.048) Average class size −0.002 (0.012) −0.001 (0.012) Constant 0.795*** (0.138) 0.777*** (0.137) Adjusted R2 0.724 0.725 N 3,267 3,267
Note: *p < .05; **p < .01; ***p < .001.
339Journal of Public Administration Research and Theory, 2016, Vol. 26, No. 2
Ta b
le 5
. A
lt e rn
a ti
v e D
e p
e n
d e n
t V a ri
a b
le s
P ar
en t
Sa ti
sf ac
ti o n
A tt
en d an
ce P ro
gr es
s R
ep o rt
Sc h
o o
l V io
le n
ce T
ea ch
er T
u rn
o ve
r
b (
SE )
b (
SE )
b (
SE )
b (
SE )
b (
SE )
M an
ag em
en t
0 .3
6 3 * *
* (
0 .0
1 8 )
0 .4
9 2 * * * (
0 .0
7 2 )
0 .1
7 6 * * * (
0 .0
1 9
) −
0 .1
8 6
* *
* (
0 .0
1 9
) −
0 .0
9 5
* *
* (
0 .0
1 9
) G
o al
s (r
es id
u al
m ea
su re
) 0 .5
4 4 * * * (
0 .0
4 6 )
1 .8
4 6 * * * (
0 .1
9 6 )
0 .3
8 7 * * * (
0 .0
5 0
) −
0 .2
4 8
* *
* (
0 .0
4 8
) 0
.1 5
5 *
* (
0 .0
5 5
) T
ea ch
er t
u rn
o ve
r − 0 .0
2 9 (
0 .0
1 5 )
− 0 .1
2 2 * (
0 .0
5 4 )
− 0 .0
8 2 * * * (
0 .0
1 7
) 0
.0 7
0 *
* *
( 0
.0 1
7 )
A m
er ic
an /A
la sk
an n
at iv
e − 0 .0
2 2 (
0 .0
3 2 )
− 0 .3
2 1 * * (
0 .1
0 4 )
− 0 .0
6 6 * (
0 .0
2 9
) −
0 .0
0 3
( 0
.0 3
2 )
− 0
.0 1
4 (
0 .0
3 5
) B
la ck
0 .0
0 1 (
0 .0
0 1 )
− 0 .0
1 0 (
0 .0
0 5 )
0 .0
0 2 (
0 .0
0 1
) 0
.0 0
2 (
0 .0
0 1
) 0
.0 0
7 *
* *
( 0
.0 0
1 )
H is
p an
ic 0 .0
1 0 * *
* (
0 .0
0 1 )
0 .0
0 1 (
0 .0
0 5 )
0 .0
0 6 * * * (
0 .0
0 1
) −
0 .0
0 5
* *
* (
0 .0
0 1
) 0
.0 0
6 *
* *
( 0
.0 0
1 )
A si
an − 0 .0
0 4 *
* (
0 .0
0 1 )
0 .0
3 6 * * * (
0 .0
0 6 )
0 .0
0 8 * * * (
0 .0
0 1
) −
0 .0
0 6
* *
* (
0 .0
0 1
) −
0 .0
0 1
( 0
.0 0
1 )
F em
al e
0 .0
0 7 *
* (
0 .0
0 2 )
0 .0
5 1 * * (
0 .0
1 9 )
0 .0
0 5 (
0 .0
0 4
) −
0 .0
1 0
* *
* (
0 .0
0 2
) 0
.0 0
2 (
0 .0
0 3
) L
im it
ed E
n gl
is h
− 0 .0
0 6 *
* (
0 .0
0 2 )
0 .0
0 6 (
0 .0
1 0 )
0 .0
0 2 (
0 .0
0 2
) 0
.0 0
2 (
0 .0
0 2
) 0
.0 0
2 (
0 .0
0 3
) R
ec en
t im
m ig
ra n ts
0 .0
0 2 (
0 .0
0 6 )
− 0 .0
3 9 (
0 .0
4 1 )
0 .0
0 1 (
0 .0
0 9
) −
0 .0
0 1
( 0
.0 0
7 )
− 0
.0 0
3 (
0 .0
1 0
) Sp
ec ia
l ed
u ca
ti o n
− 0 .0
1 8 * *
* (
0 .0
0 3 )
− 0 .1
4 0 * * * (
0 .0
1 6 )
− 0 .0
0 3 (
0 .0
0 3
) 0
.0 2
1 *
* *
( 0
.0 0
4 )
0 .0
1 3
* *
* (
0 .0
0 3
) T
em p o ra
ry h
o u si
n g
0 .0
0 4 (
0 .0
0 3 )
− 0 .0
1 7 (
0 .0
1 3 )
0 .0
0 2 (
0 .0
0 4
) 0
.0 0
9 *
( 0
.0 0
4 )
0 .0
0 2
( 0
.0 0
5 )
P o ve
rt y
ra te
0 .0
0 0 (
0 .0
0 1 )
− 0 .0
1 5 * (
0 .0
0 7 )
0 .0
0 2 (
0 .0
0 2
) 0
.0 0
5 *
* *
( 0
.0 0
1 )
− 0
.0 0
0 (
0 .0
0 2
) F
re e
lu n ch
− 0 .0
0 1 (
0 .0
0 1 )
− 0 .0
1 5 * * * (
0 .0
0 4 )
− 0 .0
0 1 (
0 .0
0 1
) 0
.0 0
2 (
0 .0
0 1
) 0
.0 0
1 (
0 .0
0 1
) R
ed u ce
d l u n ch
− 0 .0
0 2 (
0 .0
0 3 )
0 .0
4 6 * * * (
0 .0
1 3 )
0 .0
0 4 (
0 .0
0 3
) −
0 .0
0 4
( 0
.0 0
3 )
− 0
.0 1
2 *
* *
( 0
.0 0
3 )
Y ea
r 2 0 0 8
0 .5
3 8 * *
* (
0 .0
2 3 )
− 0 .2
5 8 * * * (
0 .0
7 1 )
0 .0
6 0 * (
0 .0
3 0
) 0
.0 4
2 (
0 .0
2 5
) 0
.1 1
5 *
* (
0 .0
3 6
) Y
ea r
2 0 0 9
0 .7
3 2 * *
* (
0 .0
2 8 )
0 .6
0 2 * * * (
0 .1
1 0 )
1 .1
4 4 * * * (
0 .0
3 7
) 0
.1 3
7 *
* *
( 0
.0 3
4 )
− 0
.0 2
6 (
0 .0
4 2
) E
le m
en ta
ry s
ch o o l
0 .3
5 7 * *
* (
0 .0
6 1 )
1 .2
4 2 * * * (
0 .2
3 0 )
− 0 .1
5 1 * (
0 .0
5 9
) −
0 .4
1 8
* *
* (
0 .0
6 8
) −
0 .3
8 3
* *
* (
0 .0
6 5
) M
id d le
s ch
o o l
− 0 .1
3 5 *
* (
0 .0
4 3 )
1 .4
9 1 * * * (
0 .2
0 4 )
0 .1
0 3 * (
0 .0
4 5
) 0
.3 0
4 *
* *
( 0
.0 4
8 )
0 .0
9 9
* (
0 .0
4 6
) H
ig h s
ch o o
l − 0 .3
8 3 * *
* (
0 .0
6 4 )
− 4 .9
9 9 * * * (
0 .3
2 5 )
− 0 .0
6 9 (
0 .0
6 7
) −
0 .3
7 6
* *
* (
0 .0
6 8
) −
0 .1
8 6
* *
( 0
.0 6
8 )
A ve
ra ge
c la
ss s
iz e
− 0 .0
8 2 * *
* (
0 .0
1 7 )
− 0 .1
4 8 (
0 .0
7 6 )
− 0 .0
2 9 (
0 .0
1 7
) −
0 .0
3 2
( 0
.0 1
9 )
− 0
.0 3
0 (
0 .0
1 9
) C
o n st
an t
− 0 .7
8 7 * *
* (
0 .1
6 6 )
9 2 .3
0 8 * * * (
1 .0
4 8 )
− 1 .0
6 9 * * * (
0 .2
2 3
) 0
.0 6
8 (
0 .1
6 7
) −
0 .5
1 6
* *
( 0
.1 8
0 )
A d ju
st ed
R 2
0 .5
8 3
0 .6
6 0
0 .4
4 2
0 .3
8 0
0 .1
5 4
N 3 ,2
6 7
3 ,2
6 7
3 ,2
6 7
3 ,2
6 7
3 ,2
6 7
N o te
: * p <
. 0 5 ; * * p <
. 0 1 ; * * * p <
. 0 0 1 .
340 Journal of Public Administration Research and Theory, 2016, Vol. 26, No. 2
(not shown), however, is not statistically significant; together they contribute no more than they do indi- vidually. Because student performance is a strong pre- dictor of parental satisfaction (see Favero and Meier 2013), it is also interesting to add student test score performance as another independent variable to this equation. When this is done, management has a posi- tive impact on parental satisfaction, over and above its impact on student performance. Clearly, good management of schools influences parental percep- tions, which are not completely driven by the results of standardized exams.
Student Attendance Table 5 also shows that internal management is posi- tively associated with student attendance and the offi- cial progress report and negatively associated with school violence and teacher turnover. The goals meas- ure has a similar set of relationships. The positive rela- tionship for attendance is intriguing but raises the issue of the process by which school management affects student attendance. Quite likely the relationship is indirect—working through teacher morale, student performance, and student safety. Similarly, the micro- foundations of the goals’ relationship is also interest- ing in that setting high standards for students appears to generate greater levels of student attendance. The actual processes behind these two relationships are beyond the scope of this study but should be the sub- ject of additional research.
Official Evaluations The progress report score represents the city’s overall assessment of each school and is largely based on a value-added measure of student progress in the last year. The Progress Report score also takes into account measures of student performance and attendance.8 Because the progress report includes these other factors and places a high priority on annual improvement, it should be sensitive to a variety of concerns aside from raw test scores. The positive relationships for manage- ment and goals hold even when controlling for student test scores (again the interaction, not shown, is not
statistically significant). These results suggest that the positive relationship is the result of the progress report assessing elements of education that are not part of test scores by considering annual progress and adjustments for value added.
School Safety School safety is a major issue based on both the com- monsense notion that children need a safe environ- ment in which to learn and also state and federal laws that focus on the issue. The New York annual School Violence Index provides an inverse measure of safety. This measure is created by taking a weighted sum of the number of violent incidents in a school and dividing this sum by the school’s total enrollment (New York State Education Department 2011).9 We created a logarithmic transformation of this measure and then standardized the values. Internal manage- ment has a strong negative relationship with the level of violence in the schools. This finding is a prelimi- nary indicator that good quality management and ambitious goals are likely to have spillover effects— that is, they are likely not merely to influence perfor- mance on the primary goal of the organization but can also influence other outcomes that both contrib- ute to overall success and are important in their own right (the interaction is again not statistically signifi- cant). Especially interesting is that a stress on high academic performance is associated with lower levels of school violence. Again, examining the microfoun- dations of this relationship is an interesting research topic.
Teacher Turnover Finally, the equation on teacher turnover is presented to illustrate the relationship between quality internal management and employee retention. This positive relationship for management is not surprising, given the strong theoretical support for a linkage between management, employee morale, and employee reten- tion. The results also suggest that the impact of inter- nal management is larger than the results shown in the prior tables because management also reduces turno- ver which in turn improves performance. The previous tables only demonstrate the impact of management independent of any influence through reduced turno- ver. At the same time, the establishment of high stand- ards increases turnover, a relationship that could result either from the additional stress that it places on teachers or a willingness of management to replace teachers who do not perform.
9 More severe incidents are given a larger weight. For example, forcible sex offenses are given a weight of 60, whereas weapons possession is given a weight of 15.
8 We made two changes to the city’s Progress Report score for our study. First, the city’s overall score also takes into account School Survey results. We recalculated the overall score using the same formula as the city except that we omitted the survey results, since our management variables are derived from the results of some of these survey questions. Second, the city evaluates elementary/middle schools and high schools separately, so we combined the two sets of scores into a single variable that reflects progress across both levels of schooling. To accomplish this, we standardized the two sets of scores and then took a weighted average of the two scores, using the number of students enrolled at each level of schooling (as indicated in city records) as the weight. We also combined elementary/middle school and high school attendance rates using a weighted average.
341Journal of Public Administration Research and Theory, 2016, Vol. 26, No. 2
Conclusion
This article has offered four contributions. First, the study has specified a set of core activities common to most theories of public management and has found that they contribute to performance. All have to do with basic managerial practices often encouraged in the literature of the field, and all have often been linked theoretically to changes in worker behavior that lead to better performance in public organizations. Setting clear and challenging goals, engendering employee trust, encouraging participation by frontline workers in decision making, and providing feedback to those workers have long been staples of the literature, if not always in practice. Even in settings constrained by public sector limitations on management, there is good reason to expect to find positive relationships of these practices with performance. This expectation is vali- dated in the analysis here and that is the key substan- tive finding of the study.
An important implication for policy makers and administrative leaders is to treat very seriously the importance of supporting this pattern—this virtu- ous circle of managerial practices—with procedures and routines that allow the managers to build and sustain such efforts. In urban school systems, incen- tives for professional advancement can be based in part upon mid-level managers’ undergoing training on these issues and their demonstrating willingness to put such practices into action. Where such patterns can be shown to be functioning smoothly, it is sensible to buffer those organizations from shocks and perturba- tions that can disrupt the operations.
Second, and importantly in a methodological sense, we use the subordinates’ assessments to measure man- agerial actions. The teachers are probably the most well-informed actors on the subject, since the manag- ers themselves can be expected to be at least somewhat unreliable in reporting how they manage. Teachers, by contrast, have less reason to report inaccurately.10 Whether the set of four related management practices are bundled together into a single overall measure of internal management, or whether the practices are kept empirically distinct, internal management is shown here to make positive contributions to school perfor- mance. The effect size of the contribution from internal management, furthermore, is substantively significant.
A practical implication for the public management research community, accordingly, is to look for more opportunities to tap managerial behavior in ways other than simply asking managers themselves about what they are doing. In particular, data from subordinates about their own managers and managerial settings can
be valuable and highly revealing. It is certainly worth- while to pay attention to managers’ perspectives, and often surveys of and interviews with managers can provide crucial information. But alternative sources of managerial data, such as those tapped in this study, should be investigated more often.
Third, we conduct a variety of statistical tests to make sure our findings are robust. For the primary educational outcome analyzed, student performance in standardized exams, substantial additional analysis confirms the core finding. The use of autoregressive specifications continues to show that internal manage- ment boosts performance. The purging of halo effects shows findings that remain quite similar. There is little doubt, accordingly, that the positive effects of manage- ment in the New York City school system are real.
Of particular interest is the role of challenging goals, or high expectations, in generating better performance. We find that both goals and the other features of inter- nal management contribute to performance, and the two in combination help more than the sum of the sep- arable contributions of each. Strong internal manage- ment enhances the value of high expectations; or high expectations leverage more contribution from strong internal management. For those interested in improv- ing performance in public education, it would seem that setting and communicating high expectations can be particularly important. Systematic attention to such goal setting in other kinds of public organizations is also worthy of serious attention. All these actions are within the control of managers, should they seek to employ them. Collectively they form a management style characterized by building effective teams to deal with organizational problems. It is a decentralized style that seeks to rely on the strengths of subordinates.
Fourth, the work examines a broad range of objec- tives for public schools. The internal management practices examined here are positively associated with a wide variety of outcomes, including student attendance and parental satisfaction with school performance. They also help schools achieve a vari- ety of additional valued outputs and outcomes, even some—like reducing violence in schools—that would appear to be only slightly related to the core educa- tional function. Although the correlations among all these goals are modest at best, internal management helps to achieve every one of them. In contemporary school systems, which are often beset by a wide variety of challenges and distractions, this finding is of par- ticular interest.
The patterns reported here are encouraging with regard to the importance of some traditionally empha- sized public administrative practices. Although they are restricted to one large public school system, one can expect similar patterns in other school systems. The
10 Teachers might have an incentive to report favorably on their principal, but the halo corrections that are applied should adjust for this.
342 Journal of Public Administration Research and Theory, 2016, Vol. 26, No. 2
findings may also hold in other highly professional- ized organizations, where frontline workers are highly educated and regularly engage in discretionary actions. This possibility should be on the agenda of research- ers interested in the relationship between management and performance.
This research demonstrates that mid-level public managers in the New York City school system are not helpless or irrelevant in the educational process. They and their managerial practices clearly shape edu- cational outcomes. Students of the traditional litera- ture on public administration would not be surprised at these findings, but many often-repeated ideas and claims in numerous fields of learning are eventually shown to be false. Examining the validity of such key ideas as those examined here remains important.
Supplementary Material
Supplementary material is available at the Journal of Public Administration Research and Theory online (www.jpart.oxfordjournals.org).
References Agranoff, Robert. 2007. Managing within networks: Adding value to public
organizations. Washington, DC: Georgetown Univ. Press. Andrews, Rhys, George A. Boyne, Kenneth J. Meier, Laurence J. O’Toole,
Jr., and Richard M. Walker. 2010. Wakeup call: Strategic manage- ment, network alarms and performance. Public Administration Review 70:731–41.
Argyris, Chris. 1957. Personality and organization: The conflict between sys- tem and individual. New York, NY: Harper.
Avellañeda, Claudia N. 2009. Municipal performance: Does mayoral quality matter? Public Administration Research and Theory 19:285–312.
Barnard, Chester I. 1938. The functions of the executive. Cambridge, MA: Harvard Univ. Press.
Barzelay, Michael, and Babak J. Armajani. 1992. Breaking through bureau- cracy: A new vision for managing in government. Berkeley (CA): Univ. of California Press.
Boyne, George A., Kenneth J. Meier, Laurence J. O’Toole, Jr., and Richard M. Walker, eds. 2006. Public service performance: Perspectives on measure- ment and management. Cambridge, UK: Cambridge Univ. Press.
Boyne, George A., and Richard M. Walker. 2006. Strategy content and pub- lic service organizations. Journal of Public Administration Research and Theory 14:231–52.
Chun, Young Han, and Hal G. Rainey. 2006. Goal ambiguity in U.S. federal agencies. Journal of Public Administration Research and Theory 15:1–30.
Cooper, William H. 1981. Ubiquitous halo. Psychological Bulletin 90:218–44. Costrell, Robert M. 1994. A simple model of educational standards. American
Economic Review 84:956–71. Darling-Hammond, Linda, and Deborah Loewenberg Ball. 1998. Teaching
for high standards: What policymakers need to know and be able to do. Philadelphia, PA: National Commission on Teaching and America’s Future.
Deci, Edward L., Robert J. Vallerand, Luc G. Pelletier, and Richard M. Ryan. 1991. Motivation and education: The self-determination perspective. Educational Psychologist 26:325–46.
Dirks, Kurt T., and Donald L. Ferrin. 2002. Trust in leadership: Meta-analytic findings and implications for research and practice. Journal of Applied Psychology 87:611–28.
Driscoll, James W. 1978. Trust and participation in organizational decision making as predictors of satisfaction. Academy of Management Journal 21:44–56.
Dull, Matthew. 2009. Results-model reform leadership: Questions of cred- ible commitment. Journal of Public Administration Research and Theory 19:255–84.
Favero, Nathan, and Kenneth J. Meier. 2013. Evaluating urban public schools: Parents, teachers and state assessments. Public Administration Review 73:401–12.
Feeney, Mary K., and Hal G. Rainey. 2010. Personnel flexibility and red tape in public and nonprofit organizations. Journal of Public Administration Research and Theory 20:801–26.
Gagné, Marylène, and Edward L. Deci. 2005. Self-determination theory and work motivation. Journal of Organizational Behavior 26:331–62.
Graber, Doris A. 2003. The power of communication: Managing information in public organizations. Washington, DC: CQ Press.
Gulick, Luther H. 1937. Notes on the theory of organization. In Papers on the science of administration, ed. Luther H. Gulick and Lyndal Urwick, 3–35. New York, NY: Institute of Public Administration.
Hess, Frederick M. 1999. Spinning wheels: The politics of urban school reform. Washington, DC: Brookings.
Hood, Christopher. 1991. A public management for all seasons. Public Administration 69:3–19.
Hrabluik, Coreen, Gary P. Latham, and Julie M. McCarthy. 2012. Does goal setting have a dark side? The relationship between perfectionism and maximum versus typical employee performance. International Public Management Journal 15:5–38.
Huselid, Mark A. 1995. The impact of human resource management practices on turnover, productivity, and corporate financial performance. Academy of Management Journal 38:635–72.
Ingraham, Patricia W., ed. 2007. In pursuit of performance: Management sys- tems in state and local government. Baltimore, MD: Johns Hopkins Univ. Press.
Ingraham, Patricia W., Philip G. Joyce, and Amy Kneedler Donahue. 2003. Government performance: Why management matters. Baltimore (MD): Johns Hopkins Univ. Press.
Jacobson, Willow S., Christine Kelleher Palus, and Cynthia J. Bowling. 2010. A woman’s touch? Gendered management and performance in state administration. Journal of Public Administration Research and Theory 20:477–504.
Kettl, Donald F. 2005. The global public management revolution, 2nd ed. Washington, DC: Brookings.
Latham, Gary P. 2007. Work motivation: History, theory, research, and prac- tice. Thousand Oaks, CA: Sage.
Latham, Gary P., Laura Borgogni, and Laura Petitta. 2008. Goal setting and performance in the public sector. International Public Management Journal 11:385–403.
Lawler, Edward E., and J. Richard Hackman. 1969. Impact of employee par- ticipation in the development of pay incentive plans: A field experiment. Journal of Applied Psychology 53:467–71.
Lipsky, Michael. 1980. Street-level bureaucracy: Dilemmas of the individual in public service. New York, NY: Russell Sage.
Locke, Edwin A., and Gary P. Latham. 1990. A theory of goal setting and task performance. Upper Saddle River, NJ: Prentice Hall.
Lynn, Laurence E., Jr., Carolyn J. Heinrich, and Carolyn J. Hill. 2001. Improving governance: A new logic for empirical research. Washington, DC: Georgetown Univ. Press.
May, Peter J., and Søren C. Winter. 2009. Politicians, managers, and street- level bureaucrats: Influences on policy implementation. Journal of Public Administration Research and Theory 19:453–76.
Maynard-Moody, Steven, and Michael Musheno. 2003. Cops, teachers, coun- selors: Stories from the front lines of public service. Ann Arbor, MI: Univ. of Michigan Press.
McCombs, Barbara L., and Jo Sue Whisler. 1997. The learner-centered class- room and school: Strategies for increasing student motivation and achieve- ment. San Francisco (CA): Jossey-Bass.
343Journal of Public Administration Research and Theory, 2016, Vol. 26, No. 2
McGregor, Douglas. 1960. The human side of enterprise. New York, NY: McGraw-Hill.
Meier, Kenneth J., and Alisa Hicklin. 2008. Employee turnover and organiza- tional performance: Testing a hypothesis from classical public administra- tion. Journal of Public Administration Research and Theory 18:573–90.
Meier, Kenneth J., and Laurence J. O’Toole, Jr. 2003. Public management and educational performance: The impact of managerial networking. Public Administration Review 63:675–85.
———. 2011. Comparing public and private management: Theoretical expec- tations. Journal of Public Administration Research and Theory 21 (Suppl 3):i283–99.
———. 2013a. I think (I am doing well), therefore I am: Assessing the validity of administrators’ self-assessments of performance. International Public Management Journal 16:1–27.
———. 2013b. Subjective organizational performance and measurement error: Common source bias and spurious relationships. Journal of Public Administration Research and Theory 23:429–56.
Miller, Gary J. 1993. Managerial dilemmas: The political economy of hierar- chy. Cambridge, UK: Cambridge Univ. Press.
Miller, Katherine I., and Peter R. Monge. 1986. Participation, satisfaction, and productivity: A meta-analytic review. Academy of Management Journal 29:727–53.
Miner, John B. 2005. Classics of organizational behavior: Essential theories of motivation and leadership. Armonk, NY: M. E. Sharpe.
Mirel, Jeffrey, and David Angus. 1994. High standards for all: The struggle for equality in the American high school curriculum, 1890–1990. American Educator 18:40–2.
Mosher, Frederick C., ed. 1967. Government reorganizations: Cases and com- mentary. New York, NY: Bobbs-Merrill.
Moynihan, Donald P. 2008. The dynamics of performance management: Constructing information and reform. Washington, DC: Georgetown Univ. Press.
New York City Department of Education. 2011a. General educa- tion students graduation requirements. http://schools.nyc.gov/NR/ rdonlyres/53FADF0D-D784-435E-8675-90E20624DAE1/0/2011GenEd. pdf (accessed November 4, 2011).
———. 2011b. Performance & accountability. http://schools.nyc.gov/ Accountability/default.htm (accessed July 22, 2011).
New York State Education Department. 2011. Violent and disruptive inci- dents—School safety & accountability. http://www.p12.nysed.gov/irs/ vadir/vadir-accountability.html (accessed July 22, 2011).
New York State Testing and Accountability Reporting Tool. 2011. New York state report cards. https://www.nystart.gov/publicweb/ (accessed July 22, 2011).
Nicholson-Crotty, Sean, and Laurence J. O’Toole, Jr. 2004. Public manage- ment and organizational performance: The case of law enforcement agen- cies. Journal of Public Administration Research and Theory 14:1–18.
Nye, Barbara, Larry V. Hedges, and Spyros Konstantopoulos. 2000. The effects of small classes on academic achievement: The results of the Tennessee class size experiment. American Educational Research Journal 37:123–51.
Ostrom, Elinor. 2005. Understanding institutional diversity. Princeton, NJ: Princeton Univ. Press.
O’Toole, Laurence J., Jr., and Kenneth J. Meier. 1999. Modeling the impact of public management: Implications of structural context. Journal of Public Administration Research and Theory 9:505–26.
———. 2009. The proverbs of new public management: Lessons from an evi- dence-based research agenda. American Review of Public Administration 39:4–22.
———. 2011. Public management: Organizations, governance, and perfor- mance. Cambridge, UK: Cambridge Univ. Press.
Ouchi, William G. 2009. The secret of TSL: The revolutionary discovery that raises school performance. New York, NY: Simon and Schuster.
Owens, Robert G., and Thomas C. Valesky. 2011. Organizational behavior in education: Leadership and school reform, 10th ed. Boston, MA: Pearson.
Pinder, Craig C. 2008. Work motivation in organizational behavior, 2nd ed. New York, NY: Psychology Press.
Pindyck, Robert S., and Daniel L. Rubinfeld. 1991. Econometric models and economic forecasting, 3rd ed. New York, NY: McGraw Hill.
Podsakoff, Philip M., and Jiing-Lih Farh. 1989. Effects of feedback sign and credibility on goal setting and task performance. Organizational Behavior and Human Decision Processes 44:45–67.
Provan, Keith G., and H. Brinton Milward. 1995. A preliminary theory of interorganizational network effectiveness: A comparative study of four community mental health systems. Administrative Science Quarterly 40:1–33.
Rainey, Hal G. 2014. Understanding and managing public organizations, 5th ed. San Francisco, CA: Jossey-Bass.
Riccucci, Norma M. 2005. How management matters: Street-level bureaucrats and welfare reform. Washington, DC: Georgetown Univ. Press.
Rodgers, Robert, and John E. Hunter. 1991. Impact of management by objectives on organizational productivity. Journal of Applied Psychology 76:322–36.
Simon, Herbert A. 1997. Administrative behavior, 4th ed. New York, NY: Free Press.
Thorndike, Edward L. 1920. A constant error in psychological ratings. Journal of Applied Psychology 4:25–9.
Tummers, Lars, and Victor Bekkers. 2014. Implementation, street-level bureaucracy, and the importance of discretion. Public Management Review 16:527–47.
Tyler, Tom R. 2006. Psychological perspectives on legitimacy and legitimation. Annual Review of Psychology 57:375–400.
Wilson, James Q. 1990. Bureaucracy: What government agencies do and why they do it. New York, NY: Basic Books.
Copyright of Journal of Public Administration Research & Theory is the property of Oxford University Press / USA and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
