The Autonomy Gap: Implications for Raising Student Achievement

Anthony P. Cavanna, Superintendent of Schools, The Public Schools, West Orange, NJ (
Susan Bowles Therriault, Senior Research Analyst, American Institutes for Research
Chrys C. Marcus, Research Associate, American Institutes of Research

 Cavanna Susan BowlesChrys 

 Anthony P.

 Susan Bowles

 Chrys C.


While the need for strong school leaders in American primary and secondary education is widely recognized, little effort has been made to determine whether today’s principals actually possess the authority to exercise strong leadership. Simply stated, it is not clear that school leaders have the flexibility they need to get the results demanded by state and federal accountability systems. Are their hands tied by government regulations, contract provisions, and district mandates? What do principals regard as the greatest impediments to effective leadership? How does this impact the way the school principal is viewed?

These questions became the impetus for a research study conducted by the American Institutes for Research (AIR) and the Thomas B. Fordham Foundation. The study interviewed and surveyed 33 district and charter school principals from five districts in three states. This effort focused on answering whether these principals, in an era of standards and accountability, had the autonomy they needed to impact student achievement. A pilot of the study was conducted in New Brunswick (New Jersey) Public Schools with the assistance of a group of former New York City Principals. The following is a discussion of our findings and their implications for school leaders.

How the Standards and Accountability Movement Affects
the Autonomy of Principals

Researchers, policymakers, and educators look at principalship through a framework in which the principal is the leader of a school in a loosely coupled public education system (Weick, 1976). In such a system, schools are able to act relatively autonomously. However, state and federal policies have altered the public education system so that schools, districts, and state departments of education are no longer loosely coupled. Rather, they are being tethered more tightly as the links between the accountability measures found in state and federal law are strengthened. Thus, they have a greater influence upon one another. A more intertwined relationship requires skills, resources, and capacity at each level to be reorganized and focused to function in a system where there is more interdependence on student outcome measures. This is in contrast to the autonomy at different levels of the education system that existed in previous decades. As an important facet of this change, leadership must be revisited and redefined, and this is especially true for school principals.  

Our findings suggest that:

Barriers exist but are perceived as surmountable. Principals reported that there are significant barriers to their autonomy that affect their ability to serve the children in their schools and to meet the demands of state and federal accountability systems. Some principals accepted this as part of the constraints of their positions. But others described ways to work around these barriers; most relied on the power of their personalities and strategic relationships to increase their autonomy in the areas of personnel, curriculum, and resource allocation. 

Placed between the federal, state, district, and classroom levels, the principal is in essence a middle manager. Morgan, Bacon, Bunch, Cameron, and Deis (1996, p. 360) define middle managers as “those who occupy positions between the strategic apex and the operating core of an organization.” School principals act as such between the state and district levels where strategy is laid out and the classroom where teaching and learning occurs. Principals must ensure that the demands of the larger system are being met and most importantly that students are learning. Middle managers, like principals, “play a critical role in ensuring a constancy of purpose in public organizations that have multiple sources of authority” (Morgan et al., 1996, p. 363). Indeed, principals must reconcile the tensions between federal, state, and local demand so that they make sense for the school and the classroom. The principals interviewed for this study were not ardent about changing the system of public education, rather they viewed their job as managing or muddling through the system, using relationships and informal networks.  

To lead, principals choose between bucking the system or working the system to make changes that will ultimately raise student achievement. To understand bucking versus working the system, it is important to understand different types of change. Waters, Marzano, and McNulty (2003) distinguish between first and second order change (Exhibit 1). First order change is described as an extension of the past, and second order change is a break from the past. Most educational research agrees that to meet the demands placed on states through NCLB, schools (particularly secondary schools) will have to tackle second order change processes. Not surprisingly, each school requires a different set of leadership skills.

Exhibit 1: Characteristics of First and Second Order Change

First Order Change

Second Order Change

An extension of the past

A break from the past

Within existing organizational norms

Inconsistent with prevailing organizational norms

Consistent with prevailing values

Conflicting with prevailing values

Easily learned using existing knowledge and skills

Requires new knowledge and skills to learn

Implemented with existing knowledge and skills

Requires new knowledge and skills to implement

First order change fits the style of many of the principals who participated in this study. Principals described how they had learned to function within the existing culture, norms, and boundaries of the public education system. Relationships, communication, and focus are a few of the more important characteristics of a first order change principal and were often mentioned during the interviews with principals in both district and private schools. Second order change was most frequently mentioned by charter school principals who, by the nature of the charter school movement, are trying to seek alternatives and break the mold of the traditional public education system.

School principals accept constraints as part of their job description. Principals in district-operated schools in this study accepted constraints to leadership as part of their job and as part of the public education system. This was unexpected, as we did not expect to find principals who saw themselves functioning within these challenges without any real drive to change the system. Our team identified a few possible reasons for why these principals tolerate the system as it exists. First, many of the principals have moved through the ranks of the public education system. Second, these leaders know how the system works, and more than likely were able to rise to the level of principal because they functioned highly within the system. Last, principals seemed to have ways around the system and its constraints through the positive relationships they had built with the community, teachers, district staff, and the superintendent. Most of the school leaders felt anchored in a system in which union collective bargaining agreements and district policies removed this control from their domain.

Principals are school leaders who adapt to their circumstances. In line with first order change skills and characteristics of middle managers, principals in district-operated schools attributed effective leadership to personality and building positive relationships rather than instructional leadership. The principals considered themselves realists who needed to do the best job they can with what they have, and most agreed that working in the public education system is about finite resources and working within constraints. The more experienced principals reported the least constraints, and this seemed to be rooted in the relationships they had formed inside and outside of the school and in their knowledge of the local context.

Local context makes a difference. It is clear in this study, albeit a common finding, that the local context largely dictates the ability of principals to exert influence over their school and determines their ability to lead. In districts in which principals felt supported by their superintendents, their own sense of self-efficacy to overcome barriers appeared to be greater. Dutton, Ashford, O’Neill, Hayes, and Wierba (1997) found that the role of a middle manager making changes within an organization is highly influenced by the context in which that manager works. If a change violates norms, the manager may become politically vulnerable within the organization, or if the manager has a distant relationship with upper management, then the middle manager may be discouraged from making changes or taking risks. If upper management expresses a willingness to listen to middle managers and if the overall organizational culture is supportive, then middle managers are more likely to implement change and take risks. Likewise, tenured principals expressed a sense of autonomy at a level not felt by less experienced principals. The “de facto” sense of autonomy as described by these experienced principals, was felt because they were familiar with the system and had a working relationship with teachers, district staff, and the superintendent. Not surprisingly, these principals felt the least amount of constraint.

Relationships are critical. Rather than change the system through formal mechanisms, the principals chose to work around the system through informal networks and relationships. Interviews with principals revealed that positive working relationships played an important role in school leadership. A good relationship between a principal and district staff and the superintendent improved the formal and informal communication between them. Thereby, they had a better understanding of each other and negotiated for necessary resources on formal and informal levels. Also a principal who built relationships inside the school reported a greater ability to use persuasion to lead teachers and staff.


At first, the findings of this study seem to contradict each other. First, principals believe they have the skills necessary to lead their schools. Second, many of these principals identify common barriers to leadership, such as staffing decisions and the allocation of resources. But, how do these findings fit together? Principals find themselves balancing the demands of building a school vision, maintaining a school climate, and developing professional learning communities that are conducive to teaching and learning, while taking on outside pressures of reporting requirements and demands of the district and state. Many principals described their job in terms of “juggling” many responsibilities. Indeed, principals pointed out that they feel responsible for buffering their staff from sometimes volatile external demands and policies to maximize the potential of the learning environment at their schools. With this understanding, it is important to reexamine the way in which we understand school leaders and to make appropriate adjustments so they can make optimal decisions for improving student achievement.

First, policymakers, researchers, and educators need to alter their perceptions of the principal as an autonomous leader of the school. It seems that many of the frustrations expressed by principals throughout this study are similar to those of middle managers in the corporate sector. Thus, it would serve policymakers, researchers, and educators well to alter their perceptions of principalship from that of an autonomous leader to that of a middle manager.

Second, as accountability measures have centered on student achievement and school indicators have become the measure for assessing the performance of districts and state education agencies, a stronger sense of co-dependence among the levels of the public education system has also grown. This places more scrutiny on schools and requires principals to possess exemplary and strategic relationship-building skills. Managing these relationships outside of school can increase a leader’s sense of empowerment to make critical decisions aimed at student improvement in the school. Indeed, in testimony from secondary school principals, it is clear that the better they manage these relationships, the more autonomy they muster to focus school efforts on raising student achievement. This suggests that maybe it is time to recognize that informal networks and relationship building are an inherent part of being an effective school leader. It seems that these skills are often overlooked and avoided because they are not easy to teach to future principals. However, it is clear that they are the skills that gave the more experienced principals in this study a stronger sense of self-efficacy.

Last, the difference between first and second order change must be explicit. If principals are expected to choose between the two to fulfill the mandates of state and federal accountability policies, then principals must be trained to understand these decisions and districts must create an environment in which this is acceptable. The weight of accountability indicators is placed squarely on the shoulders of principals, and pressures from districts and states add to that weight. However, we cannot expect principals to function as anything but middle managers, making incremental changes and working the system, if maintaining the system is rewarded. Second order change can only come about when educators and administrators at the state, district, and school levels understand, encourage, and reward such behavior.


Bolman, L., & Deal, T. (2002). Reframing the path to school leadership: A guide for teachers and principals. Thousand Oaks, CA: Corwin Press.

Cavanna, A., Baran, V., & O’Cummings, M. (2007). Research to practice: Effective schools are led by instructional leaders who are supported by service-oriented districts (Expert Brief for the State of New Mexico Adequacy Study). Washington, DC: American Institutes for Research.

Cavanna, A., Olchefske, J., & Fleischman, S. (2006, May). Challenging the status quo: Research matters: The potential of the portfolio approach. Educational Leadership, 63(8), 89–91.

Dutton, J., Ashford, S., O’Neill, R., Hayes, E., & Wierba, E. (1997). Reading the wind: How middle managers assess the context for selling issues to top managers. Strategic Management Journal, 18(5), 407–425.

Hallinger, P., Bickman, L., & Davis, K. (1996). School context, principal leadership and student reading achievement. Elementary School Journal, 96(5), 527–549.

Leithwood, K., & Jantzi, D. (1999). The relative effects of principal and teacher sources of leadership on student engagement with school. Educational Administration Quarterly, 35(5), 679–706.

Liethwood, K., Louis, K. S., Anderson, S., & Walhstrom, K. (2004). Review of research: How leadership influences student learning. Minneapolis, MN: Center for Applied Research and Educational Improvement at the University of Minnesota.

Morgan, D., Bacon, K., Bunch, R., Cameron, C., & Deis, R. (1996). What middle managers do in local government: Stewardship of the public trust ant the limits of reinventing government. Public Administration Review 56(4), 359–366.

Waters, J., Marzano, R., & McNulty, B. (2003). Balanced leadership: What 30 years of research tells us about the effect of leadership on student achievement. Aurora, CO: Mid-continent Research for Education and Learning. Retrieved from January 15, 2007, from

Weick, K. (1976) Educational organizations as loosely coupled systems. Administrative Science Quarterly 21(1), 1–19.