TOWARDS AN OBJECTIVE EVALUATION OF TEACHER PERFORMANCE Readers are free to copy, display, and distribute this article, as long as the work is attributed to the author(s) and Education Policy Analysis Archives, it is distributed for non- commercial purposes only, and no alteration or transformation is made in the work. More details of this Creative Commons license are available at http://creativecommons.org/licenses/by-nc-nd/2.5/. All other uses must be approved by the author(s) or EPAA. EPAA is published jointly by the Mary Lou Fulton College of Education at Arizona State University and the College of Education at the University of South Florida. Articles are indexed by H.W. Wilson & Co. Write commentary at http://epaa.info/wordpress, and send errata notes to Sherman Dorn (epaa-editor@shermandorn.com). EDUCATION POLICY ANALYSIS ARCHIVES A peer-reviewed scholarly journal Editor: Sherman Dorn College of Education University of South Florida Volume 15 Number 5 March 14, 2007 ISSN 1068–2341 Responding to the Education Reform Agenda A Study of School Superintendents’ Instructional Leadership Paul V. Bredeson University of Wisconsin at Madison Brad W. Kose University of Illinois at Urbana-Champaign Citation: Bredeson, P. V. & Kose, B. W. (2007). Responding to the education reform agenda: A study of school superintendents’ instructional leadership. Education Policy Analysis Archives, 15(5). Retrieved [date] from http://epaa.asu.edu/epaa/v15n5/. Abstract Education reforms have affected schools and the educators who work in them. Using state-wide survey data from 1993 and 2003, this study examines how the work of school superintendents has been affected over a ten-year period by these reform initiatives, especially increased demands for accountability. The general message from our data is that superintendents are interested in curriculum and instruction and believe these are important tasks, but the daily realities of their work often subvert even the most committed professional. Further, the data indicate that superintendents may be able to use external accountability mechanisms as levers to move the internal accountability systems to support improved teaching and learning. Keywords: education reform; superintendent; instructional leadership; accountability Education Policy Analysis Archives Vol. 15 No. 5 2 Respondiendo a la agenda de reformas educativas: Un estudio acerca del liderazgo educativo de los superintendentes de escuelas Resumen Reformas educativas han afectado escuelas y educadores que trabajan en ellas. Utilizando datos de encuestas estatales entre 1993-2003 este estudio examina como el trabajo de los superintendentes de escuelas fue afectado durante ese período de diez años por esas reformas educativas, especialmente por las demandas de desempeño responsable (“accountability”). El resultado general de nuestros datos, muestra que los superintendentes están interesados en los temas de currículo y enseñanza y creen que esas son tareas importantes, pero la realidad cotidiana de sus trabajos, frecuentemente subvierte hasta los profesionales más comprometidos. Además, nuestros resultados indican que los superintendentes podrían utilizar mecanismos de desempeño responsable para estimular y mejorar sistemas internos de desempeño y evaluación de la enseñanza y el aprendizaje. Palabras clave: reforma educativa, os superintendentes, liderazgo educativo, desempeño responsible. The role of the school superintendent in the United States has undergone significant changes over the past century. Once considered to be the instructional leader and teacher of teachers, more recently the discourse on the work of superintendents has shifted to politics and collaboration focused on excellence and educational. The work of superintendents has increasingly become defined by their responses to the complexities and challenges of political pressures and conflicting interests (local, state, and national), unpredictable school finances, standards-based reform, and greater demands for accountability of increased student performance through state and federal legislation such as No Child Left Behind (2001) (Farkas, Johnson, Duffett, & Foleno, 2001; Feuerstein & Dietrich, 2003; Lecker, 2002; Sherman & Grogan, 2003). Superintendents have been challenged to take on these larger responsibilities despite issues raised with the legitimacy and purpose of standards (Apple, 2001), the ethical nature of superintendents’ responses to the achievement gap (Sherman & Grogan, 2003), the feasibility of heightened superintendent role expectations without a substantial reconceptualization of superintendents’ training and authority, school board responsibilities (Fuller et al., 2003), and changes in the broader social, cultural, and political context in which schools and districts operate (Cuban, 2004). In particular, superintendents have been expected to create the district and school conditions to improve curricular, instructional, and assessment practices toward improved student learning and outcomes, which we define as instructional leadership (e.g., Rallis, Tedder, Lachman, & Elmore, 2006). Theoretically, we assume that whether one is speaking of direct, indirect, distributed, or other ways of asserting instructional leadership, increased attention and responsibility in superintendent instructional leadership is a desirable response to educational reform and accountability. However, recent large-scale empirical evidence that studies whether and how superintendents’ respond to educational reform and accountability in practice remains limited. Additionally, no known study has systematically examined how superintendents’ work has changed over the past decade within this context of reform. The purpose of this article is to present findings from two investigations of the instructional leadership roles of school superintendents in the United States. Specifically, the article examines the work of school superintendents within the context of recent policy initiatives from 1994 to 2003 at Responding to the Education Reform Agenda 3 local, state, and national levels and their impact on the ways in which superintendents define their primary work, allocate their time, and respond to increased role and policy expectations for enhanced student learning outcomes. Two major questions guided the research. What do superintendents’ priorities and work indicate about their response to educational reform? and What internal and external accountability factors influence superintendents’ response to reform? Using questionnaire and interview data from two survey investigations conducted in 1994 and 2003, we organize the paper around four key findings. First, the work of superintendents has intensified over the past decade resulting in role overload that often mitigates the impact of superintendents’ greater awareness of and responsiveness to increased demands for accountability of student learning outcomes. Second, the intersection of external policy mandates and personal/professional interests provides an important setting for understanding superintendents’ responsiveness to new demands for accountability. Third, external policy initiatives focused on student learning outcomes have helped to legitimize individual superintendent’s personal/professional interest in leadership responsibilities in the areas of curriculum, instruction, and assessment practices, especially for newly appointed superintendents and for female administrators. Given demands for accountability for student learning, this legitimacy within policy and practice arenas suggests the possibility of redefining the primary work of superintendents as educational leaders. Finally, our data indicate that too much emphasis on external mechanisms of accountability, often highly prescriptive in nature, threatens to undermine personal/professional interest as well as other important internal mechanisms of accountability. In the final section of the paper, we discuss the implications of these findings for the work of superintendents, for policy, and for future research. Theoretical Framework and Background Two theoretical frameworks—role theory and educational accountability—informed our investigation. Because superintendents hold a unique and pivotal leadership position in school districts, we were especially interested in how their work would provide a window into the journey of policy initiatives through various stages of implementation and organizational adjustment. Firestone (1989) argued that an examination of school district leadership would help policymakers understand whether reforms were viewed as opportunities or constraints. Louis (1999) and Goldring, Crowson, Laird, and Berk (2003) reported that the discourse on policy implementation and change in schools surrounds the role of school leaders. While we understand the limitations in examining policy migration to local districts/schools through only one educator role, using superintendents’ self-reports of their work in our research provided clear parameters for investigation as well as wide representation across all public school districts in one large Midwestern state. Our investigation also represents a direct response to Stevenson and Schiller’s (1999) call for more theoretically grounded studies that explore the complex linkages between state policies and school practices, which for this paper entails the work of school superintendents. One way to understand the work of school superintendents is through the lens of role theory. Role theory is a conceptual framework that describes the nature of complex and dynamic social interactions. These include role sending, role reception, role enactment, and perceptions of role enactment among interdependent groups/actors who seek to influence each other’s work (sets of expected behaviors for a given role) within a particular system (Katz & Kahn, 1966). Thus, superintendents come to understand their roles as educational leaders through a social learning process of role making (self-understanding of role) and role taking around their leadership role, the Education Policy Analysis Archives Vol. 15 No. 5 4 product of influences and expectations of others—e.g., school board members, principals, teachers, community members, and policy makers (Hart & Bredeson, 1996). Roles within organizations are rarely stagnant. Brunner, Grogan, and Bjork (2002) describe seven discursive stages that characterize shifts in the role of school superintendents. Sumbramani and Henderson (1996) describe how change in role behavior often occurs in iterative cycles of influence and responses to those influences in organizations. We hypothesize that the last decade of educational reform policy initiatives, whether locally or externally driven, were likely to have influenced the expectations various role holders had for superintendents especially in such areas as curriculum development, instructional leadership, and assessment of student learning outcomes leading to role transitions. As Petersen (1999) reported, there is an emerging “[n]ew and somewhat different leadership role for the district superintendent in the core technologies of curriculum and instruction” (p. 17). Elsewhere, we have documented superintendents’ responses to a decade of reform (Bredeson, Johansson, & Kose, 2004; Bredeson & Kose, 2005). Transition leadership (Goldring, Crowson, Laird, & Berk, 2003) suggests that it is important to examine the substantive shifts and role transitions in superintendents’ work. As superintendents negotiate the terrain between internal and external accountability systems within a volatile educational reform environment, there are transitions in role and work priorities due to demands for greater accountability for student learning outcomes. Anchored in role transition theory, superintendents’ responses to demands for greater accountability for student learning outcomes provide a window into the impact of educational reform policy on their work. We will look for evidence of role transition, “the process in which an individual changes from one set of expected role-related behaviors in a social system to another set” (Bredeson, 1993, p. 44). Accountability is another framework that provides both empirical and analytic lenses for an investigation of superintendents’ work and the policy environment. “One of the most important features of the new accountability is that it shifts governance from a system of local political accountability to state administrative accountability” (Timar, 2003, p. 178). Various educational players, including superintendents, over time will slowly and through a process of mutual adaptation adapt and adjust patterns of individual and organizational behavior to fit the new accountability policy environment (Timar, 2003). More recently federal accountability measures, most notably in ESEA (No Child Left Behind, 2001), have intensified demands for superintendents to attend to assessment and student learning outcomes in their districts. Richard Elmore describes the impact of this new model of educational accountability on the evolving role of school superintendents over the past decade (quoted in Choy, 2003): The biggest change has occurred with the introduction of performance-based accountability—the direct measurement of student performance, disaggregated by school and by type of student, and the use of that data to make judgments about how well schools perform. There are many problems with such accountability systems, the biggest may be the profound change in the work of leading and managing school systems. Many superintendents haven’t realized this change, some have acknowledged the change and grapple with how to accept it, and a few have embraced the knowledge necessary to operate in this new environment. (p. 1) Though accountability has always been part of professional educators’ reality, traditional standards of accountability focused on inputs and adequacy. More recently, the shift has been to outcomes, especially student achievement (Adams & Kirst, 1999). Highlighting the work of other scholars, these authors describe six types of accountability systems—bureaucratic, legal, professional, political, moral, and market—differing from each other in terms of mutual relationships among participants, expectations, mechanisms for implementation, and incentives. Responding to the Education Reform Agenda 5 They go on to describe the relationship between the external policy and the internal- school/district orientations on accountability. Internal systems of accountability have been described as the building blocks for organizational capacity (Newman, King, & Rigdon, 1997). The relationship between internal and external accountability systems raises important questions. For example, how do superintendents negotiate the tension between existing internal and external mechanisms of accountability within their schools and communities? To what degree are external policy demands to increase student learning outcomes already being achieved through such internal mechanisms for accountability as professionalism, responsiveness to parents, and personal ethical standards? In what way(s), if any, might external mechanisms and incentives for accountability undermine the internal mechanisms necessary at the local level for successfully meeting external policy goals? A common complaint often heard when superintendents describe their work is that there is too much to be done, with new mandates and policies piling up on administrative plates that are already overloaded (Glass, Bjork, & Brunner, 2000). For example, superintendents suggest that district governance and local politics take a substantial amount of their time and energy and serve as primary obstacles to improving student performance (Farkas et al., 2001; Fuller et al., 2003). Another conflict for superintendents is the reported discrepancy between what superintendents say is important and what they actually spend their time doing (Bredeson, 1996). A primary example of this frustration involves the core technology of education, instruction, curriculum development, and assessment of student learning outcomes. Murphy (1989) noted that superintendents spend the majority of time on issues not directly related to curriculum and instruction. He concluded that instructional leadership at the district level was managed more by default than by design. Bredeson (1996) and Bredeson & Johansson (1998) also reported that superintendents ranked curriculum development as a top priority, but then correspondingly ranked it low in terms of how they actually spent their time. The literature contains various suggestions for superintendents to be more directly involved in curriculum and instruction (Murphy & Hallinger, 1988; Petersen, 1999) and current educational reform priorities (Bjork, 1990; Hord, 1990; Kowalski, 1998; Latham & Holloway, 1999; Wimpleberg, 1988). Notwithstanding the growing body of research on the work of superintendents, there is little documentation of how work tasks and time devoted to other administrative responsibilities can be reallocated so that superintendents can focus on instructional leadership. Based on the tension suggested by the dynamic and sometimes conflicting sets of role expectations for superintendents, administrators’ self-reports on their work and how it is affected by various factors, including policies initiatives, provide a window for understanding the relationships among leadership and role change, education reform policies, and educational outcomes. Method The paper examines two sets of questionnaire data as well as structured interview data. The first survey data set was collected in 1993–1994, and the second was collected in 2003. All public school district superintendents (426) in a large Midwestern state received a six-page written survey. The respective response rates for surveys returned were 82.1% and 75.2%. The original questionnaire (Bredeson, 1996) was modified only slightly for data collection in 2003, allowing direct comparisons of superintendents’ self-reports of their work as well an analysis of key factors that influenced their work and response to major reform initiatives over the past decade. Education Policy Analysis Archives Vol. 15 No. 5 6 Data Collection In each survey year, we first mailed a survey to all superintendents in the fall term. After one month, approximately 64% of superintendents returned completed surveys. A second mailing was sent to all non-respondents. Though non-responses are always a consideration in survey research, we believe the return rates of 82.1% (1994) and 75.2% (2003) are satisfactory return rates to have confidence in our findings. To determine whether or not superintendents who returned the survey were a distinct population from non-respondent superintendents, we examined the demographic profile of all superintendents in the state and compared it to our sample. There were no meaningful differences in the profiles of these two groups. Survey instrument. To address the major research questions, a six-page written questionnaire was sent to the all superintendents. The questionnaire included four types of survey items: demographic data from respondents (e.g., age, training, experience, district enrollment, and prior professional experiences); twelve open-ended questions that asked superintendents to describe their work in terms of priorities, importance, time spent, tasks for which they were accountable in their performance evaluations, and the most important things they did in the area of instructional leadership and curriculum development; Likert-scaled items in which superintendents were asked to rate such items as involvement in curriculum development, current level of satisfaction with their work in curriculum and instructional leadership, and the expectations others have for superintendents in curriculum and instruction; and rank-order items in which respondents were asked to rank order major work tasks by importance, time spent on each task, and the major factors that affected their involvement in curriculum development and instructional leadership. Structured interviews. After collecting and analyzing the two sets of questionnaire data, we conducted 12 structured interviews with practicing superintendents in 2003. The survey data yielded a rich overview of the landscape of superintendent work and their responses to various reform initiatives. Yet we wanted fuller and more personal narratives of the challenges and opportunities within the context of small rural schools, small city and suburban school districts, and large urban districts. Each interview lasted between 60 and 90 minutes. All interviews were recorded, transcribed, and prepared for analysis. Data Analysis All survey data were coded and prepared for analysis. Responses with numeric data—Likert ratings, rank order data, and nominal data—were coded for analysis using SPSS (Statistical Package for Social Sciences, 2004). Descriptive statistics, correlations, and tests of mean differences were used to analyze the data. All data reported statistically significant are at a p value of .05 or less. To help ensure independence in statistical testing when comparing 2003 and 1994 data, only superintendents with fewer than 10 years experience for the 2003 survey (N=191) were included in the comparisons which used statistical tests compared to 1994. We recognize the trade-off by making this change. It is possible that the sample is less representative of the population. However, the assumption of independence required for statistical inference outweighed using both full samples for these comparisons. Open-ended responses were transcribed into electronic format and then stored and analyzed within the electronic qualitative software program QSR NVivo 2.2 (1999–2002). We began by using a constant comparative method (Bogdan & Biklen, 1998) to create categories/codes for all responses. Using a randomly selected set of surveys, codes were developed for each of the 13 open- ended survey items. Next, we tested the utility and exhaustiveness of these codes by using a second Responding to the Education Reform Agenda 7 randomly selected set of surveys. Two researchers used the codes developed and refined in the open coding phase to code data on three selected items on 20 randomly selected surveys. Agreement between the two researchers independently coding these data ranged from .733 to .950. Any differences in coding between the two researchers were highlighted and discussed in order to increase the reliability of the final coding. Once all open-ended data were coded, we collaboratively constructed themes within each of the 13 open-ended survey items before looking for themes that stretched across the survey. Similarly, we developed codes for interview transcripts by using a constant comparative method and constructed themes from these codes. We then compared survey and interview themes to check for correspondence and to look for more detailed elaboration of key findings and themes. Findings We organize the findings around our two guiding research questions. We present findings that describe how superintendents think about, define, and carry out their work within two sections: The first section highlights the changing descriptive profile of superintendents. The second section delineates superintendents’ espoused instructional leadership priorities and practices in 2003 and a comparison of noteworthy 1994 and 2003 instructional leadership data. Next we describe what these descriptive data tell us about how superintendents have adapted to the new realities of accountability and what various adjustments and adaptations mean in terms of the links between internal and external demands for accountability. A Descriptive Profile of Superintendents Age. Our data indicate there were differences in the average of age of superintendents when we compare 1994 to 2003 data. These differences reflect current national trends in the superintendency. The median age in years for all superintendents was up from 50.0 in 1994 to 53.0, while the distribution extended from 33 to 66 in 1994 to a range of 30 to 72 for 2003. This increase mirrors a 1999–2000 national survey (Gates et al., 2003) where the median age of superintendents rose to 52.5, the highest median age ever recorded in this survey. Gender, experience, highest academic degree. While the superintendency continues to have a substantially higher proportion of males in 2003 (83.5%, N=264), the percentage of females (16.5%, N=52) is more than double the percentage reported in the 1994 survey (6.5%). This growing female representation corresponds to national data that indicate 13 percent of superintendents are female, double the percentage from 1992 (Gates et al., 2003). In our study, male superintendents had more years experience in the superintendency (with a mean of mean 10.1 years) compared to women (6.6 years). The 1994 survey had similar statistical differences. However, our data indicate this gap may be closing with women in 1994 at 4.9 average years of experience in the superintendency and men at 10.4 average years. For the 2003 survey, the mean number of years for all superintendents in their current position was 5.4, with no significant gender differences. On average superintendents held three professional positions (roles included teacher, assistant principal, principal, counselor, curriculum director, business manager, special education administrator, and/or assistant superintendent) prior to the superintendency. There were no significant gender differences. Of those surveyed, 57% of these administrators had only worked in one school district, while 17.1% had worked in one other district, 14.9% had worked in two other districts, and 6.6% had worked in three or more districts prior to their current appointment. The Education Policy Analysis Archives Vol. 15 No. 5 8 cumulative percentage of 95.6% of these first four tiers stands in sharp contrast to the image of the superintendency as a revolving door reported in major urban districts (Council of the Great City Schools, 2001). Given the general size of districts in the state, with few major urban districts, the tenure of these superintendents aligns with national survey data reported by Cooper, Fusarelli, and Carella (2000). In terms of highest academic degree, 20.6% obtained a master’s degree (15.4% of female superintendents with a masters, 21.7% male), 41.3% a specialist degree (28.8% female, 43.7% male), and 38.1% a doctoral degree (55.8% female, 34.6% male). Women were statistically more likely to obtain their degree more recently on average than men (9.9 and 13.4 mean years since the highest degree, respectively). Table 1 summarizes these differences. Table 1 Superintendent Descriptive Data by Sex Proportion Median age Mean experience Highest academic degree 2003 Sex 1994 2003 1994 2003 1994 2003 Masters Specialist PhD Female 6.5% 16.5% 47.0 51.0 4.9 6.6 15.4% 28.8% 55.8% Male 93.5% 83.5% 51.0 53.0 10.4 10.1 21.7% 43.7% 34.6% District size. The average pupil enrollment of districts in which 2003 superintendents worked was 1834, an increase of nearly 50% from 1252 in the 1994 data. For purposes of analysis, four district size categories that were created. This distribution was unsurprisingly skewed toward the smaller three: 19% of superintendents worked in districts of fewer than 500 students, 29.7% ranged from 500 and 999, 44.3% varied from 1000 to 4999, and 7.0% were school districts with more than 5000 students. Superintendents’ Priorities and Work We organize various data clusters to illuminate superintendents’ priorities and work, particularly as they relate to their roles in curriculum, instruction, and evaluating student learning outcomes. These data include comparisons between the 1994 and 2003 data as well as independent 2003 data where comparison data were unavailable or irrelevant. 2003 open-ended response. In an open-ended question in 2003, we asked superintendents to describe the three most important things they do. They reported budget and finance (18.3%), communications (16.3%), personnel administration (14.8%), and curriculum development (12.5%) as their most important tasks. 1994 and 2003 rankings of task importance and time allocation. The open-ended responses corroborate well with 1994 and 2003 superintendent rankings of major task areas by importance (see Table 2). Although the task area of evaluation of student learning outcomes was introduced in 2003 (this category was not included in the 1994 survey), the mean rankings for both task importance and time spent are remarkably similar for both surveys. The 2003 independent sample ranked all task areas and time spent categories in the same order as the total 2003 sample. Responding to the Education Reform Agenda 9 Table 2 Superintendent Ranking of Task Area Importance and Time Spent Mean Rankings in 1994* Mean Rankings in 2003 Task area Importance Time Importance Time Budget and school finance 3.23 2.66 3.58 3.32 Planning/goals formulation 3.99 5.67 4.39 5.63 Community/public relations 4.34 4.56 4.76 4.88 Curriculum and instructional leadership 4.36 5.67 4.60 5.78 Personnel administration 4.50 3.50 5.01 3.55 Evaluation of student learning outcomes** — — 5.58 6.91 Professional growth and staff development 5.75 6.72 6.04 6.72 School board relations/training 5.88 5.75 6.51 6.41 Legal/political issues 6.37 5.36 6.77 5.08 Facilities management 6.50 5.07 7.51 6.57 * 1 = most important ** Evaluation of student learning was not an option in 1994 Attending to the budget and school finances (annual operating funds) dominates the top ranking for each survey in both importance and time, while curriculum and instructional leadership leapt from the fourth highest ranking in 1994 to the third highest ranking in 2003 in importance. With the exception of time spent on facilities management, all categories stayed the same or moved only one ranking spot from 1994 to 2003. The disparities between task importance and time allocation related to curriculum development and instructional leadership, evaluation of student learning outcomes, and professional growth and staff development in both surveys are perhaps most striking. In 1994 curriculum and instructional leadership fell from fourth in importance to seventh in time spent; in 2003, the same category respectively dropped from third to sixth. In 2003, evaluation of student learning outcomes (which was not requested in 1994), plummeted from sixth in importance to tenth in time spent. In 1994, professional growth dropped from sixth in importance to ninth in time spent; in 2003, it fell from seventh to ninth, respectively. A comparison of the two data sets reveals a few statistically significant differences, but not in curriculum and instructional leadership. The 2003 superintendents from the independent sample ranked the following as less important than superintendents in 1994: facilities management, personnel administration, and school board relations/training. The 2003 independent-sample superintendents reported spending less time in facilities management and school board relations and training than superintendents in 1994. For this question involving task importance and time spent, we also found statistical differences within each survey for gender and district size. For 1994, only the importance of facilities management was found to be more important for men than women. In the complete 2003 sample, our analysis revealed significant differences between female and male superintendents in areas of importance. Males ranked budget and finance, facilities management, personnel management, community/public relations, and legal-political issues higher than did female administrators (p=.03). Females ranked curriculum development and instructional leadership and evaluation of student learning outcomes higher than their male counterparts. These data may indicate a shift in priorities Education Policy Analysis Archives Vol. 15 No. 5 10 based on gender, perhaps most notably in the areas of teaching and learning. However, we did not find statistical differences in time spent between females and males in 1994 or in 2003. It appears that regardless of differences in how male and female superintendents judge the important of a task, the time for more important priorities such as planning and curriculum development and instructional leadership is subsumed by time spent on less important priorities such as legal and political issues and facilities management. A comparison of the four district sizes by using the Kruskal-Wallace test of K-samples also reflected differences in the ways superintendents described their most important responsibilities. Using a chi-square test of statistical significance, one can conclude that larger-district superintendents placed greater importance on planning and goals formulation, professional development, and evaluation of student learning. Facilities management and budget and finances dominated superintendents’ work in small districts. Similarly, we found significant differences in time spent. In both 1994 and 2003, smaller-district superintendents ranked budget and finances and facilities management higher while ranking time spent on planning and goals formulation as less important. For superintendents in larger districts, school board training/relations required more time than in smaller districts. These data may indicate that superintendents in smaller districts need to attend more to budget and facilities management than superintendents in larger districts due to their multiple roles, whereas administrators in larger districts may be able to delegate more of the budget responsibilities and concentrate on visioning and planning to central office staff. Broadly speaking, these rankings indicate that superintendents view curriculum and instructional leadership (and to a certain degree evaluation of student learning outcomes) as important, but other responsibilities—especially legal and political issues—consume the time they would like to devote to instructional leadership. This issue particularly seems to be the case for women in 2003. One superintendent captured the tension created by competing and often times overwhelming work demands: “With the new development of the expectations with the testing and the data analysis, and so forth, it became apparent after 3 or 4 years that this was a killer job trying to do both.” Evaluation of student learning. In 2003, we asked respondents to describe the most important things they did as superintendents in the area of evaluation of student learning in schools. Of 622 items listed in response to this question, data analysis was the most frequently cited task (in 29.9% of responses). Management of curriculum and instruction (21.1%), providing support (15.9%), goal setting (11.7%), communication of student learning outcomes (6.9%), delegating and setting expectations (6.9%) and professional development (6.8%) were among the most important ways that superintendents involved themselves in the evaluation of student learning in their districts. Note that data analysis, which is clearly linked to external accountability influences, was over 40% more frequent than the next most common response of monitoring curriculum and instruction. There was not a matching item to this query in the 1994 survey. However, the open-ended responses to questions regarding the most important things superintendents did in areas of instructional leadership and curriculum development offered no references to data analysis in 1994. Perception of involvement with curriculum and instruction. Superintendents in both surveys were asked whether they believed their involvement in curriculum and instructional leadership had increased, stayed the same, or decreased over the past five years. The descriptive statistics in Table 3 seem to indicate an increased involvement in curriculum and instructional leadership from 1994 to 2003. Our 1994 data indicate respective percentages of 34.1%, 40.7% and 25.2%. The 2003 independent data (that controls for superintendents in both surveys) demonstrated a statistically significant increase in curriculum and instruction (48.4%) while 22.5% remained the same and 29.5% decreased in involvement. Responding to the Education Reform Agenda 11 Table 3 Superintendent Change in Curriculum and Instructional Leadership over Prior 5 Years Survey Year Involvement Increased Stayed the Same Decreased 1994 34.1% 40.7% 25.2% 2003 48.4% 22.5% 29.1% The next six data areas all involve 2003 data which shed additional light on how superintendents have responded to reform and the extent to which they value and prioritize improved curriculum, instruction, and student learning. These include responsibility to support principals’ instructional leadership, hiring priorities, analysis of preparation, and budget allocations. Responsibility to support principals’ instructional leadership. When asked to rate how important it is for the superintendent to be responsible for principals’ instructional leadership and school improvement (1=not important and 5=very important), both female and male superintendents gave very high ratings. Female respondents’ mean rating (4.92) was higher than males’ mean rating (4.72). There were no significant differences in ratings based on prior experience in other administrative roles, years of administrative experience, or district size. Satisfaction with curriculum and instruction. We also asked respondents to rank their satisfaction with curriculum and instructional leadership (from 1 as very satisfied to 5 as very dissatisfied). Both males (mean of 2.71) and females (2.31) reported moderate satisfaction which suggests they see room for improvement. Female superintendents reported greater satisfaction than male superintendents. For this same survey, larger district superintendents were found to be more satisfied in this leadership area than their counterparts, which may be explained by their ability to delegate more of this responsibility. Satisfaction was not related to the number of years experience as superintendents. Self-comparison to other superintendents. When asked whether or not they thought they were typical superintendents in terms of levels of involvement in curriculum development and instructional leadership, 87.2% of the female superintendents did not believe they were typical while only 44.8% of male superintendents reported they were atypical. When we examined whether years of experience influenced their sense of typicality, we found that superintendents with fewer than 10 years of experience as superintendents believed they were atypical versus those with 10 or more years as superintendents. There was no significant relationship between school district size and superintendents’ responses to being typical in their involvement in curriculum and instructional issues. Hiring priorities. In 2003, we also asked superintendents a series of questions about their work and the hiring of administrative assistants. If you could hire an assistant to whom you would delegate specific administrative tasks, what type of assistant would you hire and why? This was followed by, How would hiring this assistant change your work? Based on their 327 responses, five major types of administrative assistants were identified. These included a general assistant (in 32.4% of responses), curriculum director (23.2%), business manager (18.3%), human resources/personnel administrator (10.4%), and an assistant for communications and public relations (7.0%). Of those who provided a rationale for any type of assistant (N=170), 40.0% wanted general relief while 39.4% explicitly wrote the purpose was to directly or indirectly enhance curriculum and instruction. Data generated from both sets of questions indicate that superintendents wanted someone to share their work load and also saw the improvement of curriculum and instruction as a top priority. Preparation. Superintendents’ opinions on their preparation provide a different window into understanding the current demands and priorities of their work. When asked what area(s) they would like to have been better prepared as a superintendent, the top three categories were budget Education Policy Analysis Archives Vol. 15 No. 5 12 (20.5%), political/legal (18.1%) and curriculum and instruction/evaluation of student learning outcomes (14.4%) Whereas curriculum, instruction, and evaluation of student learning continue to be top priorities, the political/legal emphasis corroborates with superintendents spending more time on an issue they ranked as less important (ranked ninth in importance and fourth in time in the previously mentioned ranking of task areas and time spent). These data indicate that, along with other data such as what type of assistant superintendents would hire, the daily demands and various roles that superintendents play may dissuade more involvement in curriculum and instruction. Changes in budget. Our data also provide evidence that many superintendents and districts increased their financial allocations to curriculum, instruction, and student learning. We asked superintendents, In what ways, if any, has your local school district budget changed to match new priorities in curriculum development and student learning outcomes? Surprisingly, in a context of decreasing educational funding, 68.6% of the responses (N=293) indicated an increase in budget spending for improving curriculum and instruction or student learning while 26.6% of the budgets remained the same, and 4.8% cut improvements to curriculum and instruction or student learning. One superintendent wrote, “Because of budget issues, it has been difficult to free up necessary funds. However, we have placed a large emphasis on student achievement which has resulted in greater emphasis on staff development.” For superintendents in districts where curriculum and instruction remained steady or was cut, 82.6% implied that they would have spent more money on curriculum development and instructional improvement if it were available. In fact, 14 respondents stated that they reluctantly cut curriculum and instruction initiatives; this does not include respondents who complained of budget cuts writ large. While these statistics continue to indicate that superintendents prioritized curriculum and instruction to enhance student learning, it appears that budget shortages prevented superintendents from using resources toward curriculum and instruction from improving. Several superintendent responses ranged from expressions of anger and frustration to an attitude of “doing the best we can” because of inadequate funding and the accompanying sacrifices that were made. Two quotes represent this range: Superintendent 1: In my district the fiscal pressures are tremendous and have resulted in many changes that have impacted student learning. I’m very angered by this. No money…no personnel…no programs. It is that simple! Superintendent 2: We have gone from a half-time curriculum director to a full- time person. We have resisted budget cutting in those areas that are critical to teachers engaging students in the learning process. (This state) must commit to greater funding of K-12 education and seek out other tax revenues besides property taxes to fund K-12 education. Whether examining 2003 data, comparisons of 1994 and 2003 data, disaggregated data (especially by gender), or qualitative or quantitative data, this section as a whole suggests that superintendents have responded to reform and valued improved curriculum development, instructional leadership, and evaluation of student learning. However, open-ended questions as well as quantitative data suggest that other areas of responsibility and role overload prevent superintendents from matching task time allocation with task importance. The next section examines the relationships between internal and external accountability factors that influenced superintendents’ role expectations, motivations and response to reform. Responding to the Education Reform Agenda 13 Internal and External Accountability Influences While the previous section indicates that superintendents have responded to reform, this section highlights internal and external accountability factors that influence superintendents’ response to reform. As noted earlier, Adams and Kirst (1999) delineated internal and external forms of educational accountability. Moral, professional, and political areas of accountability characterize internal accountability. External accountability is largely driven by market, bureaucratic, and legal mechanisms. Instructional leadership expectation ratings. Role expectations, closely aligned with professional accountability, contribute to superintendents’ understanding of their responsibilities and actual work in the area of curriculum development and instructional leadership. When we asked superintendents to rate the expectations others had for them in this area (from 1 for low expectations to 5 for high expectations), superintendents rated school boards and principals expectations highest (see Table 4). When we compared female and male superintendents (N=312), there were statistically significant differences in their understandings of what others expected of them. Female superintendents (n=52) reported higher expectations from their school board, teachers, students, and parents than their male counterparts (n=260). Table 4 Superintendents’ Rating of Others’ Expectations in Curriculum Development and Instructional Leadership Role Senders Mean Rating of Expectations School Board 4.25 Principals 4.15 Teachers 3.99 Parents 3.83 Community 3.71 Students 2.81 School board annual performance criteria. At the same time, open-ended superintendent responses indicate that these school board high expectations in curriculum and instruction were not top priorities with the criteria they use to evaluate superintendents’ annual performance. Superintendents’ annual performance review is closely tied to professional, bureaucratic and legal accountability (both internal and external accountability). School boards expected superintendents to be accountable for budget and finance (27.4%), personnel administration (19.6%), communications (19.1%), evaluation of student learning outcomes (10.6%), leadership (8.9%), general management of the district (5.2%), curriculum and instruction (4.8%), and planning and goals formulation (4.6%). These differences suggest that school boards have yet to see planning and goals formulation and curriculum/instruction as critical responsibilities of the superintendent. Even when evaluation of student learning and curriculum and instructional leadership are combined (15.4 %), they fall short of each of the top three categories. Qualifications for recruiting and hiring superintendents. Another open-ended question provided indirect insight into superintendents’ priorities as well as thinking about internal and external accountability: What kinds of qualifications (expertise) should school boards be looking for when they recruit and hire superintendents? Combined categories of expertise in curriculum and instruction and student evaluation/achievement (13.1%) continue to show the congruence between external and internal accountability (which again are heavily influenced by both state mandates and Education Policy Analysis Archives Vol. 15 No. 5 14 personal interest). However, a large portion of the responses (52%) focused on internal accountability. Internal accountability responses included: general leadership, which is associated with professional accountability (20.2% of responses, including delegating leadership, visionary, change agent); character and disposition, which are tied to moral accountability (11.0%); communication/public relations, which are aligned with political accountability (11.8%); and “people skills,” which are associated with professional and political accountability (9.0%). In contrast, external accountability categories of budget (market accountability, 10.5%) and political/legal expertise (legal accountability, 3.9%) were surprisingly low given their dominating presence in previous data analysis. These data are consistent with one superintendent’s analysis of the relationship between these systems of accountability: “So I think it is internal and external factors that interact with one another. I would attribute more of the growth to internal factors frankly than the external factors because I think there are going to be a wide range of behaviors across [the state] in response to academic expectations and decreasing resources.” Factors that influenced curriculum and instruction involvement. We also wanted to understand which factors, either internal or external, influenced the choices superintendents made regarding their involvement in curriculum development and instructional leadership. Table 5 displays the mean rankings of factors that affected their work (from 1 as most important to 8 as least important). Table 5 Factors Influencing Superintendent Involvement in Curriculum and Instruction Factor Mean Rank Ranking as Most Influential State curriculum and testing mandates 3.08 31.3% Personal interest 3.17 33.2% Local district priorities 3.20 19.0% School reform issues 4.18 7.1% School board interest 4.32 2.7% Community interest 5.69 1.0% Staff request for superintendent involvement 5.86 2.7% Controversial issues 6.42 2.7% Superintendents gave state curriculum/testing mandates the most influential mean rank and second highest percentage of most influential top category votes (31.3%). Although there were no statistically significant differences by gender or district size, superintendents who had ten or more years of experience as superintendents were less likely to rank state curriculum/testing mandates as the number one motivator compared to their younger counterparts. The latter statistic suggests that superintendents newer to the field may pay greater attention to external accountability demands. Personal interest, clearly associated with moral (internal) accountability, had the highest percentage of number one rankings (33.2%) and second highest mean rank as a factor that influenced superintendents’ involvement with curriculum and instruction. Female superintendents were more likely to rank personal interest as the most important influence (mean=2.18) than males (mean=3.33). The percentage of personal interest (internal accountability) as the number one ranking was nearly equivalent to the combined percentages of the remaining six categories (local district priorities, school reform issues, school board interest, community interest, staff requests for superintendent involvement, and controversial issues). Therefore, the most frequent responses of state mandates and personal interest suggest that both external and internal accountability were motivators Responding to the Education Reform Agenda 15 for enhancing curriculum development and instructional leadership. The power of these two factors is further revealed when considering that school board interest and staff requests were ranked very low in terms of influence. Yet superintendents rated school boards, principals, and teachers highest in terms of expectations that others had for them in the areas of curriculum and instruction. Several interview responses corroborated the congruence between internal and external accountability influences. One respondent noted, “The one most huge thing that has caused the most change is the No Child Left Behind Act. This accountability has caused me to get closer to the instructional side of the district.” Another superintendent said that changes in his work over the past five years influenced by educational reform “gave me an opportunity to do something I really believe in and often times don’t have enough time to practice or opportunities to practice.” Similarly, another superintendent said, “The external factors as not necessarily good news, but they are reminders of some of the kinds of things we should be doing anyway. So you don’t really see them as enemies but as opportunities.” Superintendents also provided broad perspectives on the relationship between internal and external accountability in interviews. One noted that he is “more accountable in recent years than in prior years, but what is interesting, though, is that we thought we were accountable back then.” Another said, I think for the most part what we’re seeing nationwide is just more accountability, whether it be from the federal or the state level, and so there’s the focus on school improvement is coming as a result of No Child Left Behind. I think there’s some good parts to it, there’s some atrocious parts to it. The reality is we’ve got to deal with it. And, you know we’ve got to dot our i’s and cross our t’s, and I guess I try to be an optimist and say, “How can this make our school better?” And, I think we’re findings some ways to make our school better just by planning a little bit more rigorous curriculum and trying to make sure the kids don’t fall through the cracks. Finally, a different superintendent implied the need to keep internal and external forms of accountability in healthy balance: “2003 is a very, very different thing and just trying to stay connected to [internal and external forms of accountability]. Because really, if you don’t stay connected to that, I think you as a policy maker, a decision maker, I think you can really screw up a system, screw up a community really, really bad[ly].” Discussion We now turn to a discussion of our findings and their implications for theory, practice and educational policy. Two research questions guided this article. What do superintendents’ priorities and work indicate about their response to educational reform? and What internal and external accountability factors influence superintendents’ response to reform? Our data indicate that superintendents are responding to new external demands for accountability through various curriculum and instruction priorities, budget increases to support these priorities, increased attention to data analysis, and priorities in hiring assistants to support their work. Some of this shift appears to be explained by an increase in female superintendents who have strong interest in these areas. However, when we examine their work within the areas of curriculum development and instructional leadership, there is a mismatch between the realities of their daily work and recent reform policy mandates and initiatives. Role overload, a persistent condition of superintendents’ work, forces both superintendents and policymakers to confront a role dilemma. The general message from our data is that superintendents are interested in curriculum and Education Policy Analysis Archives Vol. 15 No. 5 16 instruction and believe these are important tasks, but the daily realities of their work often subvert even the most committed professional. Consistent with other research (Bredeson, 1996; Bredeson & Johansson, 1998; Farkas et al., 2001; Fuller et al., 2003), areas such as political and legal issues mitigate against superintendents’ ability to spend optimal time to enhance the core technology of teaching and learning in districts. We can only conclude that such unyielding role demands preclude significant reallocation of time and re-conceptualization of their primary work. In terms of the second research question, superintendents indicated that they respond to both internal and external accountability factors. For example, both state mandates and personal interest were top factors that influenced their involvement in curriculum and instruction. As open- ended survey data as well as interviews suggest, superintendents may be able to use external accountability as a lever to move the internal system to support improved teaching and learning. However, there appear to be some mismatches between and within internal and external accountability mechanisms. Although superintendents rated school boards with high expectations for their instructional leadership, they indicated that school boards held them much more accountable for budget, personnel, and communications than evaluation of student learning, curriculum, and instruction in their annual performance review. This may explain why school boards were rated low in their influence on curriculum and instruction. Perhaps most importantly, while many superintendents reported increases in budget allocation to curriculum and instruction, a substantial portion reported that these distributions stayed the same or declined despite desires to increase resources to areas that enhance student learning. From a policy perspective, it appears that external accountability mechanisms such as state mandates have helped legitimize superintendents’ efforts to pay greater attention to improving areas of teaching and learning. However, many superintendents and districts were unable to shift additional resources into these areas because of budget constraints, which are heavily associated with local taxes. Again, as one superintendent said, “(This state) must commit to greater funding of K-12 education and seek out other tax revenues besides property taxes to fund K-12 education.” Implications for Theory: Role, Accountability, and Transition Leadership Within complex systems, superintendents assume formal administrative roles that are shaped by personal preferences (personalization), organizational expectations (socialization), and a variety of other role messages, both internal and external. The shifts and differences reflected in the 1994 and 2003 data sets are anchored in this constellation of personal interests, expectations, and such forces as the education reform environment buttressed by mandated standards and accountability initiatives. The data from the 2003 survey support the notion that superintendents are not only aware of external accountability shifts to student achievement, but have prioritized and responded to this change in accountability, and thus are shifting from a role that has historically been defined by management and community/public relations to one that focuses on improving student learning. In addition, when one compares 1994 survey responses to 2003 survey responses, there is substantial evidence that there are higher expectations for superintendents to take action in the areas of instructional leadership, curriculum development, and student learning. The most telling indicator is the prominence of data analysis in the area of student learning outcomes in 2003. In 1994, data analysis was not mentioned as a critical task area for superintendents. We suspect that like accountability, accountability may have always been a part of the superintendent’s work, but not with an emphasis on learning outcomes. The convergence of superintendent’s leadership role demands within schools and districts coupled with increased demands for accountability suggests the contemporary superintendency is Responding to the Education Reform Agenda 17 likely to remain a dynamic and malleable leadership role shaped by both internal and external environments. Our findings resonate with those of Golding et al. (2003), providing further support for the notion of transition leadership. Superintendents need a particular set of transition leadership skills for negotiating the sometime chaotic and often uneven terrain of internal and external accountability systems within volatile and often times uncertain educational reform environments. Lastly, it is not clear that the external policy environment understands existing internal systems of accountability sustained by norms of professional practice, moral and ethical principles, and the micro-politics of local schools and districts. The finding that personal interest is such a powerful contributor to superintendent involvement and commitment to student learning and school improvement reveals an important gap in theories of accountability. We find little mention or evidence in the current literature which argues for the importance of recognizing and building on personal interests/intrinsic motivators which sustain internal accountability systems. Thus, it is important for policymakers to understand how such external accountability forces (bureaucratic, legal, and market mechanisms) interact with incentive systems designed to affect how superintendents and other school leaders align their primary work with policy goals. State and federal policies and incentive systems that build on already strong internal accountability mechanisms, such as personal interest, are more likely to lead to school improvement and enhanced student learning outcomes than an over reliance on external mechanisms. Our findings echo those of Wills and Peterson (1992): “One focus of attention for state policymakers may be the knowledge that flexibility in reaching a common set of outcomes is preferable to a rigid program that does not take into account variability in district resources and responses by district leaders” (p. 259). Thus, an important contribution from this study to accountability theories and systems is an understanding of the complementary possibilities between internal and external mechanisms for accountability mediated through the leadership of school superintendents. Our data also support Newman, King, and Rigdon’s position (1997) that absent an understanding of the connections between internal and external accountability systems, it is likely that the demands of the external accountability system may undermine local internal mechanisms of accountability thereby weakening their capacity for improvement and sustained development. Implications for the Work of Superintendents Our findings also have important implications for the daily work of superintendents. First, the shift from educational inputs as a system of accountability to educational outputs (student learning outcomes) has already affected where superintendents pay attention, even if other administrative tasks often frustrate their attempts to focus on instruction, curriculum, and student learning outcomes. Data analysis retreats, workshops, and seminars are now a significant part of the professional discourse and professional development agendas of superintendents. As mentioned, the 1994 survey data provide no evidence that data analysis was a priority either in rhetoric or practice. Thus, it seems reasonable to conclude that educational policies focusing on measures of student learning outcomes have moved educational outcomes from lofty goal statements to critical information in data-driven decision making. The prominence of student learning outcomes suggests new expectations from school boards, parents, teachers, and principals that shape superintendents’ work. However, we find a significant gap between these expectations and what school boards formally hold superintendents accountable for in their annual performance reviews. Though we have little data in this study to support it, we suspect that school boards have already moved in this direction by hiring superintendents with greater expertise and commitment to curriculum, instruction, and skills to assess student learning outcomes. Education Policy Analysis Archives Vol. 15 No. 5 18 This change raises another important issue. Given that budget and finance, personnel administration, and public relations will also continue to be critical to successful leadership, it is likely that school boards and other educational stakeholders will have to reassess their current role expectations for superintendents. Especially in small districts, the superintendent’s role is unlikely to be reshaped until role overload and priorities for leadership are realigned with more realistic expectations for the superintendent’s work. Heroes, martyrs, and over-achievers have always been part of the landscape of the superintendency. However, we should not expect these archetypes to define the superintendent’s leadership role. As transition leaders, superintendents will need to help their school boards, professional and support staffs, parents, and communities redefine role expectations for themselves as well as other professionals in their districts. With student learning outcomes as an animating force for role realignment, it is important at the outset to recognize that there is no one way to accomplish this uniformly across some 14,500 school districts in the United States. District size, local context and history, and various organizational and structural differences will likely account for very different redefinitions of superintendent work. We are not arguing that superintendents will become test specialists or curriculum directors. Rather, their energies and priorities for leadership will be shaped by new expectations anchored in educational outputs, especially indicators of student learning for all students. We argue that fundamental redefinition of superintendents’ work may affect the recruitment, selection, and career trajectories of school leaders in general and superintendents in particular. With teaching, learning, and student outcomes as key organizers for superintendent work, candidate personal interest, expertise, and experience in areas of instruction, curriculum and assessment may well generate a new pool of potential school leaders. Our data indicate that female superintendents have increased their representation in the administrative ranks and that they are more experienced, trained, and interested in curriculum, instruction, and student learning. While character, people skills, and general leadership skills will remain important attributes for aspiring superintendents, accountability demands centered on student learning outcomes legitimizes personal interest in, expertise, and experiences richly informed by careers much more focused on instruction and its outcomes. Together with findings from other empirical studies on the work of school superintendents, the findings from these data across a decade provide evidence of shifts in the discourse around the roles and work of school superintendents. Superintendents are savvy political actors who acknowledged educational changes swirling about them and the new demands for accountability and student learning anchored in state and national policy initiatives. However, our data indicate that these shifts are more real in the rhetoric describing their work than it is in the daily realities that drive the work routines of these school leaders. There continue to be discrepancies between what superintendents say is both important and personally interesting to them as leaders and how they are able to redefine or reallocate their time in ways that would substantially change the nature of their work in the area of curriculum development and instructional leadership. Kowalski (1998) argues that the evaluation of superintendents is in need of reform. In this vein, it seems reasonable to argue that attention needs to be given to the alignment of external mandates and accountability with the evaluation of superintendents’ performance. Our studies provide little evidence that the evaluation criteria used by school boards and municipalities to evaluate the performance of superintendents has changed in ways that reflect new accountability demands for curriculum reform and school/student outcomes. However, significant realignment of evaluation criteria to hold superintendent more accountable as educational leaders is unrealistic absent consideration of the current demands placed upon them and the realities of local context. Responding to the Education Reform Agenda 19 Implications for Future Research Gender disparities revealed in our data suggest an important area for further research and discussion. These data suggest that female and male superintendents have a tendency to understand their roles as leaders differently, especially around curriculum development, instructional improvement, and student learning. Females had greater personal interests in these issues, believed others had higher expectations for them, rated and ranked instruction and curriculum development higher than their male colleagues, and were more satisfied with their involvement in curriculum and instruction. Notwithstanding these differences, these self-report data indicate that female superintendents were not able to distinguish themselves from male superintendents in terms of the amount of time they were able to dedicate to curriculum development, instructional leadership, and student learning outcomes. Given these data, we hypothesize that superintendents with more expertise, experience, and commitment to curriculum, instruction, and student learning are more likely to be advocates for sustained and coherent focus on the core mission of schools—teaching and learning. Implications for Policy Despite misgivings and at times harsh criticism of NCLB, external accountability efforts seem to have both shaken up and legitimized a focus on learning within the internal system. For example, our data indicate that student outcomes are being used as a tool to change and align the internal system. At the same time, our data indicate that superintendents have responded to policy reform initiatives, but awareness does not equal implementation because of role overload. Solutions may include better superintendent preparation programs to efficiently handle managerial, legal, and political responsibilities and challenges; more resources that enable superintendents to hire general assistance and/or curriculum and instructional assistance; and a reconceptualization of superintendent roles and responsibilities. It is also paramount to consider how funding directly or indirectly shapes superintendents’ work. In states with declining resources, often exacerbated by declining enrollments, policy options to further reducing school budgets make superintendents even more powerless to enhance student learning outcomes. Again, superintendents’ ability to be responsive to external accountability is already overwhelmed by internal expectations and the reality of their work. Because external demands are always implemented at the local level, policy makers should consider how budget constraints and concerns affect superintendents’ efforts in curriculum and instructional leadership. Since superintendents report budgeting as a top priority across the board, it seems logical that increased funding would to some degree allow them to concentrate on the core mission of schools. This particularly seems to be the case for smaller district superintendents who chose providing material and resources as their most important critical priority more often than superintendents in larger districts. At the same time, policymakers are not blameless for creating mandates which require time, resources, and energy (e.g., filling out documentation) that in some ways detracts from superintendents’ instructional leadership capacity. As the previous statements imply, policy is often not sensitive to contextual variables such as gender, district size, rural or urban communities. Policies should be developed with flexibility to allow the local level to adapt and negotiate its implementation according to its context and operating mechanisms. While we are arguing for a good fit between external and internal accountability, we find conceptions of complete alignment problematic. Policy should still appreciate and acknowledge the unique purposes and qualities of internal accountability. Moral and professional realities, Education Policy Analysis Archives Vol. 15 No. 5 20 motivations, and goals are not easily amenable to broad indicators. We hope policy makers and those involved in the work of superintendents consider the democratic purposes of schooling. While standardized test scores are important indicators of student achievement, educational accountability must strive for more than increased test scores. Schools of education, educational leaders, and policy makers must together examine how all students are prepared to acquire the skills, knowledge and dispositions necessary for future employment, successful relationships and democratic participation in an increasingly diverse and complex society. Schools do need to be accountable to the public, but there is more than one way to be accountable. Certainly, we need to examine critically the extent to which what is pragmatically and efficiently measured is an indicator of what is most important. It seems time to strike a more rationale and responsive balance between educational outcomes and educational inputs and resources. Responding to the Education Reform Agenda 21 References Adams, J. E., & Kirst, M. W. (1999). New demands and concepts for educational accountability: Striving for results in an era of excellence. In J. Murphy & K. S. Louis (Eds.), Handbook of Research on Educational Administration (2nd ed., pp. 463–489). San Francisco: Jossey- Bass. Apple, M. W. (2001). Educating the "right" way: markets, standards, God, and inequality. New York: RoutledgeFalmer. Bjork, L. (1990, April). Effective schools—effective superintendents: The emerging instructional leadership role. Paper presented at the Annual Meeting of the American Educational Research Association, Boston, MA. Bogdan, R. C., & Biklen, S. (1998). Qualitative research for education: An introduction to theory and methods. Needham Heights, MA: Allyn & Bacon. Bredeson, P. V. (1993). Letting go of outlived professional identities: A study of role transition and role strain for principals in restructured schools. Educational Administration Quarterly, 29(1), 34–68. Bredeson, P. V. (1996). Superintendents’ roles in curriculum development and instructional leadership: Instructional visionaries, collaborators, supporters, and delegators. Journal of School Leadership, 6(3), 243–264. Bredeson, P. V., & Johansson, O. (1998). Leadership for learning: A study of instructional leadership roles of superintendents in Sweden and Wisconsin. In R. Muth & M. Martin (Eds.), Toward the Year 2000: Leadership for Quality Schools. Lancaster, PA: Technomic Publishing, Inc. Bredeson, P. V., Johansson, O., & Kose, B. W. (2004, April). A comparative study of superintendents in Sweden and the United States: Responses to Educational Reform Initiatives. Paper presented at the Annual Meeting of the American Educational Research Association, San Diego, CA. Bredeson, P. V., & Kose, B. W. (2005, April). School superintendents as instructional leaders: Responses to a decade of reform 1994–2004. Paper presented at the Annual Meeting of the American Educational Research Association, Montreal, Canada. Brunner, C. C., Grogan, M., & Bjork, L. (2002). Shifts in the discourse defining the superintendency: Historical and current foundations of the position. In J. Murphy (Ed.), The challenge of school leadership: Redefining leadership for the 21st century (pp. 211–238). Chicago: National Society for the Study of Education. Choy, C. P. (2003). Reconnecting superintendents to instruction: An interview with Anrig Professor Richard Elmore [Electronic Version]. Harvard Graduate School of Education Education Policy Analysis Archives Vol. 15 No. 5 22 News. Retrieved October 10, 2006, from http://www.gse.harvard.edu/news/features/elmore12012003.html. Cooper, B. S., Fusarelli, L. D., & Carella, V. A. (2000). Career crisis in the school superintendency? The results of a national survey. Arlington, Virginia: American Association of School Administrators. Cuban, L. (2004). Meeting challenges in urban schools. Educational Leadership, 61(7), 64–69. Farkas, S., Johnson, J., Duffett, A., & Foleno, T. (2001). Trying to stay ahead of the game: Superintendents and principals talk about school leadership: Public Agenda. Feuerstein, A., & Dietrich, J. A. (2003). State standards in the local context: A survey of school board members and superintendents. Educational Policy, 17(2), 237–256. Firestone, W. A. (1989). Using reform: Conceptualizing district initiatives. Educational Evaluation and Policy Analysis, 11(2), 151–164. Fuller, H. L., Campbell, C., Celio, M. B., Harvey, J., Immerwahr, J., & Winger, A. (2003). An impossible job? The view from the urban superintendent's chair. Seattle: University of Washington, Center on Reinventing Public Education. Gates, S. M., Ringel, J., Santibanez, L., Chung, C., & Ross, K. (2003). Who is leading our schools? An overview of school administrators and their careers. Santa Monica, CA: Rand. Glass, T., Bjork, L., & Brunner, C. (2000). The study of the American School Superintendency: A look at the superintendent of education in the new millennium. Arlington, VA: American Association of School Administrators. Goldring, E., Crowson, R., Laird, D., & Berk, R. (2003). Transition leadership in a shifting policy environment. Educational Evaluation and Policy Analysis, 25(4), 473–488. Hart, A. W., & Bredeson, P. V. (1996). The principalship: A theory of professional learning and practice. In Principals' influence on student and school outcomes. New York City, NY: McGraw-Hill. Hord, S. M. (1990). Images of superintendents’ leadership for learning. Austin ,TX: Texas Association of School Administrators. Katz, D., & Kahn, R. L. (1966). The social psychology of organizations. New York: John Wiley & Sons. Kowalski, T. (1998). Critiquing the CEO. The American School Board Journal, 185(2), 43–44. Latham, A., & Holloway, J. (1999). The profession of school superintendents: An analysis of the responsibilities and knowledge areas important for beginning school superintendents. Princeton, NJ: Educational Testing Service. http://www.gse.harvard.edu/news/features/elmore12012003.html Responding to the Education Reform Agenda 23 Lecker, A. (2002). Listening to leaders: The obstacle course. American School Board Journal, 189(2), 32–35. Louis, K. S. (1999). Rethinking school improvement. In J. Murphy & K. S. Louis (Eds.), Handbook of research on educational administration (2nd ed., pp. 45–72). San Francisco: Jossey-Bass. Murphy, J. (1989, April). The reform of school administration: Pressures and calls for change. Paper presented at the Annual Meeting of the American Educational Research Association, San Francisco, CA. Murphy, J., & Hallinger, P. (1988). Characteristics of instructionally effective school districts. Journal of Educational Research, 81(3), 175–181. Newman, F. M., King, M. B., & Rigdon, M. (1997). Accountability and school performance: Implications from restructuring schools. Harvard Educational Review, 76(1), 41–69. Petersen, G. J. (1999). Demonstrated actions of instructional leaders: An examination of five California Superintendents. Education Policy Analysis Archives, 7(18). Retrieved March 12, 2007, from http://epaa.asu.edu/v7n18/. Rallis, S., Tedder, J., Lachman, A., & Elmore, R. (2006). Superintendents in classrooms: From collegial conversation to collaborative action. Phi Delta Kappan, 87(7), 537–545. Sherman, W. H., & Grogan, M. (2003). Superintendents' responses to the achievement gap: An ethical critique. International Journal of Educational Research, 6(3), 223–237. Stevenson, D. L., & Schiller, K. S. (1999). State education policies and changing school practices: Evidence from the National Longitudinal Study of Schools, 1980–1993. American Journal of Education, 107, 261–288. Subramani, M., & Henderson, J. C. (1996). Gaps that matter: The influence of convergence in understanding on IS service quality. Paper presented at the Second AIS Americas Conference on Information Systems, Phoenix, AZ. Timar, T. B. (2003). The “new accountability” and school governance in California. Peabody Journal of Education, 78(4), 177–200. Wills, F. G., & Peterson, K. D. (1992). External pressures for reform and strategy formation at the district level: Superintendents’ interpretations for state demands. Educational Evaluation and Policy Analysis, 14(3), 241–260. Wimpleberg, R. K. (1988). Instructional leadership and ignorance: Guidelines for the new studies of district administrators. Education and Urban Society, 20(3), 302–310. Education Policy Analysis Archives Vol. 15 No. 5 24 About the Author Paul V. Bredeson University of Wisconsin-Madison Email: bredeson@education.wisc.edu Paul V. Bredeson is Professor and Chair of the Department of Educational Leadership and Policy Analysis. His research interests center on professional development and instructional leadership and the impact on the work of school leaders. Brad W. Kose University of Illinois at Urbana-Champaign Email: bkose@uiuc.edu Brad W. Kose is an assistant professor in the Department of Educational Organization and Leadership. His research interests include principal leadership for learning and leadership for social justice. He is currently a co-principal investigator of a longitudinal program of research that employs cultural capital frameworks to study secondary organizational practices that impact diverse students’ success in postsecondary education. Responding to the Education Reform Agenda 25 EDUCATION POLICY ANALYSIS ARCHIVES http://epaa.asu.edu Editor: Sherman Dorn, University of South Florida Production Assistant: Chris Murrell, Arizona State University General questions about appropriateness of topics or particular articles may be addressed to the Editor, Sherman Dorn, epaa-editor@shermandorn.com. Editorial Board Noga Admon Jessica Allen Cheryl Aman Michael W. Apple David C. Berliner Damian Betebenner Robert Bickel Robert Bifulco Anne Black Henry Braun Nick Burbules Marisa Cannata Casey Cobb Arnold Danzig Linda Darling-Hammond Chad d'Entremont John Diamond Amy Garrett Dikkers Tara Donohue Gunapala Edirisooriya Camille Farrington Gustavo Fischman Chris Frey Richard Garlikov Misty Ginicola Gene V Glass Harvey Goldstein Jake Gross Hee Kyung Hong Aimee Howley Craig B. Howley William Hunter Jaekyung Lee Benjamin Levin Jennifer Lloyd Sarah Lubienski Susan Maller Les McLean Roslyn Arlin Mickelson Heinrich Mintrop Shereeza Mohammed Michele Moses Sharon L. Nichols Sean Reardon A.G. Rud Lorrie Shepard Ben Superfine Cally Waite John Weathers Kevin Welner Ed Wiley Terrence G. Wiley Kyo Yamashiro Stuart Yeh Education Policy Analysis Archives Vol. 15 No. 5 26 Archivos Analíticos de Políticas Educativas Associate Editors Gustavo E. Fischman & Pablo Gentili Arizona State University & Universidade do Estado do Rio de Janeiro Asistentes editoriales: Rafael O. Serrano (ASU–UCA) & Lucia Terra (UBC) Hugo Aboites UAM-Xochimilco, México Armando Alcántara Santuario CESU, México Claudio Almonacid Avila UMCE, Chile Dalila Andrade de Oliveira UFMG, Brasil Alejandra Birgin FLACSO-UBA, Argentina Sigfredo Chiroque IPP, Perú Mariano Fernández Enguita Universidad de Salamanca. España Gaudêncio Frigotto UERJ, Brasil Roberto Leher UFRJ, Brasil Nilma Lino Gomes UFMG, Brasil Pia Lindquist Wong CSUS, USA María Loreto Egaña PIIE, Chile Alma Maldonado University of Arizona, USA José Felipe Martínez Fernández UCLA, USA Imanol Ordorika IIE-UNAM, México Vanilda Paiva UERJ, Brasil Miguel A. Pereyra Universidad de Granada, España Mónica Pini UNSAM, Argentina Romualdo Portella de Oliveira Universidade de São Paulo, Brasil Paula Razquin UNESCO, Francia José Ignacio Rivas Flores Universidad de Málaga, España Diana Rhoten SSRC, USA José Gimeno Sacristán Universidad de Valencia, España Daniel Schugurensky UT-OISE Canadá Susan Street CIESAS Occidente,México Nelly P. Stromquist USC, USA Daniel Suárez LPP-UBA, Argentina Antonio Teodoro Universidade Lusófona, Lisboa Jurjo Torres Santomé Universidad de la Coruña, España Lílian do Valle UERJ, Brasil http://epaa.asu.edu/epaa/board/schugur.html http://epaa.asu.edu/epaa/board/torres.html Theoretical Framework and Background Method Data Collection Data Analysis Findings A Descriptive Profile of Superintendents Superintendents’ Priorities and Work Internal and External Accountability Influences Discussion Implications for Theory: Role, Accountability, and Transition Leadership Implications for the Work of Superintendents Implications for Future Research Implications for Policy References Production Assistant: Chris Murrell, Arizona State University