What Would it Cost to Coach Every New Principal ? An Estimate Using Statewide Personnel Data

In this paper, I use Levin and McEwan’s (2001) cost feasibility approach and personnel data obtained from the Superintendent of Public Instruction to estimate the cost of providing coaching support to every newly hired principal in Washington State. Based on this descriptive analysis, I estimate that the cost to provide leadership coaching to every newly hired principal as an induction support would range between $153,000 and $845,000 per year representing a per pupil cost between $4.01 and $12.35, respectively, depending on the amount each leadership coach receives in compensation. A more conservative approach that provides coaching support only to newly hired principals working in high poverty schools would cost the state between $143,975 and $443,000 per year representing a per pupil cost between $4.20 and $12.92, respectively, depending on the amount each leadership coach receives in compensation. Implications for coaching and principal professional development are discussed within the context of policy decisions. In particular, the article considers whether coaching is a financially feasible professional development strategy and how new information about the per pupil cost of coaching might better inform ongoing policy discussions related to leadership coaching. epaa aape Education Policy Analysis Archives Vol. 22 No. 55 2


Introduction
The cost of leadership coaching is a significant issue for policymakers considering coaching as a strategy for administrator professional development.While coaching shows considerable promise as a professional development strategy (Bickman, Goldring, DeAndrade, Breda, & Godd, 2012;Grissom & Harrington, 2010;Huff, Preston, & Goldring, 2013;Simkins, 2009), policymakers often hesitate to provide this support given what they perceive to be excessive costs (Daresh, 2004).
In fact, the perceived cost of coaching is frequently cited as a potential barrier to providing coaching support on a wider scale (Daresh).These perceptions, however, may be inaccurate or framed without empirical research.For example, only one recent empirical study exists that specifically considers the cost of coaching support in a K-12 setting.This study focused on the cost born by a school district to provide instructional coaches to support classroom teachers (Knight, 2012).Other research highlighted the total cost to provide coaching support to administrators, but did not identify the per pupil costs that are often central to policy discussions (Villani, 2006).
The majority of estimates that relate to the cost of coaching for leaders are either not empirically based or come from business or other fields, which differ significantly from education (International Coaching Federation, 2012).According to one recent estimate, there may be as many as 40,000 leadership coaches worldwide, generating nearly $2 billion in annual revenue (International Coaching Federation).Dividing the total revenue by the number of coaches worldwide would suggest that coaches may be earning as much as $500,000 each.Another recent estimate suggests that coaches working in business and other fields earn as much as $3,250 per hour (The Conference Board, 2008).Both estimates are likely an alarming prospect for any policymaker considering the use of scarce public funds to support the development of leaders in K-12 schools.Yet, the estimates that currently exist are unlikely to provide policymakers with the kind of information they need to adequately consider the costs of coaching within the context of public education.Moreover, given the increasing interest in coaching for administrators, new information about the costs is needed in order to inform policy conversations.
With few exceptions (e.g., Knight, 2012), the research about the cost of coaching is primarily focused on state level programs operated by departments of education, collaboratives between multiple state agencies, or individual universities.For example, Villani (2006) reviewed five statelevel induction programs for novice administrators and provided a comprehensive assessment of the program structures, locations, and costs.Noting examples from Arkansas, Indiana, Mississippi, Ohio, and Tennessee, Villani (2006) found that these programs generally ranged in size and scope, as well as in total cost.Villani reported that the cost of the coaching programs ranged from $80,000 to more than $1 million.Cost variations appear to be closely related to the amount paid to coaches in the form of salaries or stipends, as well as total costs to operate the coaching programs.Other research on coaching has focused on coaching programs within individual schools, but not specifically on leadership coaching for school principals.Knight (2012), for example, studied instructional coaches in three schools and calculated the cost of the coaching intervention to be between $3,260 and $5,220 per teacher.Both studies were limited in that they did not present these costs in terms of per pupil expenses -which are central to policy-related discussions.Without specific consideration of the cost at a per pupil level, these studies do little to provide policymakers with the kind of empirical evidence they need to overcome political resistance to implementing new professional development in austere times, or to expanding coaching support that is perceived by many to be a costly venture.Further, neither of these studies sought to project the cost of coaching at a statewide scale, as they did not use actual data about the number of principals who may need to be supported.

Purpose of this Article
Given the absence of information regarding the cost of coaching and the implications that these limitations have in terms of policy development, in this paper, I draw upon administrative personnel data from Washington state to estimate the cost to provide one year of leadership coaching support to novice school administrators across a state education system.My estimate considers the cost to compensate leadership coaches on a per client basis who would be deployed to support novice administrators in the state's elementary and secondary schools.I selected Washington for this analysis as the state provides public access to the personnel data needed to produce the estimates, and has a relatively stable principal workforce in terms of annual turnover.The dataset used to establish the number of principals who may require coaching support has also been used in previous studies of teacher and administrator retention and mobility across the state system (Plecki, Elfers, Loeb, Zahir, & Knapp, 2005).
For this analysis, I determined the individual cost factors using research on coaching compensation and constructed the estimate using the cost feasibility approach described by Levin and McEwan (2001), which assigns costs to individual program components using market prices and empirical research to establish the cost estimates.In this paper, cost refers to the expense born by a state to provide coaching as a professional development support to school principals.Costs include both the amount each coach is compensated as well as the administrative expenditures needed to manage a coaching program.Given the limitations of the existing research about the efficacy of coaching (Crow, 2012), this paper does not attempt to assess the effectiveness of the intervention.Rather, the paper presents a descriptive analysis that characterizes the costs that Washington state would bear if it were to provide coaching support to novice school principals as a form of statesponsored induction.
The paper begins with a brief review of relevant literature and includes research regarding the cost of professional development, as well as research on leadership coaching as an induction strategy for novice administrators.I use this discussion to build a case for providing support for novice administrators, as well as to highlight the current limitations of the existing literature base.Next, I discuss the data sources and methods I used to complete the analysis.Finally, I present the cost estimates and conclude by discussing their implications for policy as it relates to leadership coaching.

Relevant Literature
Given the increasing need to provide support for school administrators, particularly novice administrators, policymakers increasingly face important questions about the cost of support.There is an abundance of research focused on professional development for classroom teachers, including analyses that consider the costs of professional development (Darling-Hammond, 1999;Rivkin, Hanushek, & Kain, 2005;Rockoff, 2004).According to one estimate provided by the National Staff Development Council, professional development for classroom teachers is provided at a much higher rate than professional development for school administrators (Caldwell, 1986).Surprisingly, there is relatively little research about professional development for school administrators.While researchers advocate that support for principals is critical, they have typically done so with little consideration of the cost of specific professional development interventions.Professional development is one form of support that states and districts can and often do provide to administrators.Researchers have studied the amount of money spent by states, districts, and schools on professional development, but most of these analyses have focused on classroom teachers (Clotfelter, Ladd, & Vigdor, 2007;Rivkin, et al., 2005).The literature indicates that districts and schools spend differing amounts of money on professional development, ranging from two to six percent of their total operating budget (Hawley-Miles, Odden, Fermanich, & Archibald, 2005;Killeen, Monk, & Plecki, 2002).Research conducted by Killeen et al. (2002), which used national data sets and adjusted costs based on geographic variations, estimated that school districts across the United States spent approximately three percent of their annual budgets or about $200 per pupil on teacher professional development.On average, they determined that the five school districts they studied spent approximately $19 million or $4,380 per teacher on professional development in the school year.Scholars indicate that professional development expenditures go beyond single line item entries in school and district budgets.They include direct and indirect costs associated with planning, supplying, delivering, and evaluating the professional development activities.Indeed, as Hawley-Miles et al. (2005) determined in a study of professional development in five urban school districts, professional development spending was managed by multiple offices and brought together multiple funding streams.
Research on spending for administrator professional development activities has received far less attention in the research literature.While researchers generally assume that administrative professional development consumes a smaller share of school district spending, how much is spent on principal professional development is unclear (Rice & Cohen, 2007).The cost to provide professional development to principals represents a significant policy consideration.Who should bear this cost, what a reasonable cost is, and how these costs should be monitored are all significant policy questions.Further, given growing empirical evidence that highlights the positive effects that principal leadership has on schools, understanding and discussing cost considerations becomes an important component of state and district discussions about instructional improvement.While researchers have worked to identify current investment levels in professional development, as previously stated, much of this research has focused on professional development spending for classroom teachers and has not considered professional development for school principals specifically.One of the few exceptions is a study conducted by Rice and Cohen (2007), which advanced a model for calculating the cost of professional development for school administrators.As Rice and Cohen asserted, policymakers … need comprehensive information that goes beyond a line item in a budget.They need detailed analyses of how much it costs to deliver and operate a principal professional development program, what types of resources are needed, and what individuals and organizations are expected to provide those resource.(p. 3) Accordingly, policymakers need information that can help them make informed decisions about the types of professional development to offer, as well as the costs and benefits of various professional development strategies.

Leadership Coaching: Definitions and Costs
Leadership coaching has been advocated as a viable strategy to support novice school leaders (Barnett & O'Mahony, 2008); yet, information about the cost to provide leadership coaching to administrators is virtually non-exist in education research.Coaching is generally characterized as a one-on-one relationship designed to support reflection and skill development (Bloom, Castagna, Moir, & Warren, 2005;Killburg, 1996).More specifically, coaching involves an individual with "experience, expertise, wisdom, and/or power who teaches, counsels, and helps less experienced or less knowledgeable personals develop professionally and personally" (Danielson & McGreal, 2000, p. 253).Daresh (2004) indicated that coaching support for novice administrators may be beneficial for at least two reasons.First, novice administrators benefit by working with an experienced colleague who can help them become familiar with the "policies, procedures, and normal practices in a school district" (p.502).Second, novice administrators benefit from the intense focus on refining their practice in response to feedback about their performance in "traditional skills associated with effective performance in administrative roles" (p.502).
While the benefits of coaching for novice administrators has been fairly well established in the educational leadership literature (Silver, Lochmiller, Copland, & Tripps, 2009;Fletcher, 2012;Rhodes, 2012), the discussion about the cost of coaching has primarily been carried out in business settings (The Conference Board, 2008).Related studies have provided evidence about the cost that companies bear when they provide coaching to middle and senior executives.A 2008 survey of coaching compensation, for example, found that executive coaches earn between $250 and $3,250 per hour (The Conference Board, 2008).As noted by The Conference Board, "The hourly rates that companies pay for coaching range from under $200 to more than $500" (p. 1).Unfortunately, these studies provide relatively few clues regarding the cost of providing coaching in education settings for school administrators, as most coaching arrangements in education do not function in the same manner nor under the same contractual terms as in for-profit companies.
Research that has provided some information about the cost of coaching has predominately focused on instructional coaching for classroom teachers.For example, Knight (2012) conducted an analysis of instructional coaching programs for classroom teachers in three schools.His analysis revealed that coaching costs between $3,260 and $5,220 per teacher.Further, Vilanni's (2006) discussion of mentoring and induction programs for novice principals is a notable exception to the dearth of research on coaching costs.Her analysis focused specifically on induction programs for novice administrators.Vilanni identified five state-funded induction programs for novice principals, including programs operating in Arkansas, Indiana, Mississippi, Ohio, and Tennessee.The programs she studied were housed in the state education agency and provided mentoring support to newly hired principals.Based on her analysis, the programs served between 25 and 150 principals per year at a cost ranging from $30,000 to $800,000.Some of the programs involved cost-sharing with the state principal's association, while others did not.The programs were created as "states are identifying the need to support new principals.These initiatives are intended to enhance the new principal's ability to promote high student achievement in their schools and increase the retention of new principals" (Villani, 2006, p. 14).Interestingly, though Villani provided an estimate of the total cost of the program, she did not provide per-administrator or per-pupil estimates about the cost of providing induction support.Other research indicates that the concerns about the cost of these programs, as well as the appearance of privileging any particular group, can often detract from policymakers' willingness to support the programs despite the relatively modest cost to start the programs initially (Daresh, 2003(Daresh, , 2004)).However, the fiscal realities of state education budgets run counter to the increasing body of research suggesting that principals have a demonstrable, though indirect, effect on student learning (Hallinger & Heck, 1996, 2009;Hallinger, 2011;Leithwood & Louis, 2012), and could also greatly benefit from coaching programs (Daresh, 2004).

Methods
I used Levin and McEwan's (2001) cost feasibility approach to complete this analysis.Levin and McEwan defined the cost feasibility approach as "the method of estimating only the costs of an alternative in order to ascertain whether or not it (the alternative) can be considered" (pp.23-24).This method was appropriate for this analysis, as the empirical evidence needed to assess the effectiveness of the professional development intervention (i.e., coaching) was not available in all but a few cases (Bickman, et al., 2012).Thus, I completed this analysis to determine: (a) whether the cost of leadership coaching support for novice school administrators exceeded other comparable supports described in the existing literature; and (b) how the costs of providing this support varied given different programmatic alternatives.Consistent with Levin and McEwan's approach, I identified individual program components (i.e., ingredients) and associated costs with them.I used existing literature and market analyses to determine individual cost components.In total, I estimated prices for various cost inputs, including: personnel, facilities, equipment, supplies, and required client inputs (e.g., release time, substitutes, transportation, etc.).The total of these individual costs were used to determine the total cost to provide the professional development intervention.
The cost estimates derived in this study rest on several assumptions about the type of induction support provided to principals, as well as the conditions under which this support is provided.Given the paucity of research about effective coaching practices, I drew from research I conducted on an existing university-based coaching program as a model (c.f., Silver, Lochmiller, Copland, & Tripps, 2009;Lochmiller & Silver, 2010;Lochmiller, 2014).A central aim of the program was to provide coaching support to each newly hired school administrator.Coaches are compensated with an annual stipend for each coachee they work with.For this analysis, I assumed that a similar coaching program could be funded by the state and would provide novice school administrators with one-on-one coaching support for a period of one year.I assumed that coaches who worked with administrators have recently retired from a school district in the state.There was no cost for the administrator to receive coaching support.The coaches who worked with the administrators received a small stipend for each administrator they worked with.The coaches were expected to work with the administrator for four to six hours per month throughout the school year, at a time and location of the administrator's choosing.In addition to their work with the administrators, the coaches were required to participate in ongoing professional development about effective coaching practices.The professional development was provided at no cost to the coach and was facilitated by paid staff who oversee the coaching program.The coaches' travel expenses to participate in the training were also reimbursed by the program.

Data
Data for this analysis came from the Washington Superintendent of Public Instruction's Office of Apportionment and Financial Services.I used data from the S-275 statewide personnel database, which provides information about each employee working in public schools in the Washington education system.The database includes the employee's demographic information (e.g., gender, race/ethnicity, age) and professional information (e.g., highest degree earned, salary, etc.).I also obtained school-level demographic information from the Superintendent's office.This information provides information about each school in the state, including its total student enrollment, racial/ethnic composition of the student body, socioeconomic status measured by the number of students eligible for free or reduced price lunch, and the number of students served by special education and bilingual education programs.

Merging Files and Completing the Analysis
Both datasets, the S-275 personnel file and the school demographic information, contained a unique district and school identification number that could be linked to other state data sources.In this case, the personnel information I extracted from the S-275 personnel database was merged with district and school-level demographic information.The final dataset included information about the individual's employment status, as well as information about their school and school district.The latter information enabled me to calculate how much each intervention cost on a per pupil basis.

Analysis
My analysis used simple descriptive statistics to present the total and per pupil costs of coaching support.I began the analysis by selecting personnel records for principals and assistant principals employed in the state in 2002/03, 2003/04, 2004/05, 2005/06, and 2006/07.I made the selection based on the employee's duty root code, which describes the person's position in the school as reported by the district on formal personnel records.I selected employees with duty root 21 through 24, as these individuals are reported by school districts in the personnel data working in the role of a school administrator (i.e., principal or assistant principal).Individuals with a duty root 21 or 23 are employed as elementary or secondary school principals.Individuals with a duty root of 22 or 24 are employed as elementary assistant principals or secondary vice principals.
Individual school principals and assistant principals had multiple lines of data associated with their employment.To create unduplicated lists of principals and assistant principals, I assigned each principal to a primary school location.I determined their primary school assignment using the assignment FTE reported in the personnel database.To determine the principal's assignment, I used the FTE reported in the database.I assigned each principal and vice principal to a primary school assignment.Any assignment with 0.5 FTE or greater was considered their primary assignment.In the event that the assignment was split equally (e.g., two buildings with 0.5 FTE each), I selected the building randomly.For principals working in multiple buildings, I assigned the principal to the school with the majority of their assignment FTE.
Next, I linked the records obtained for each of the years with two prior years of data to determine whether the principal was newly hired into their role.This step was necessary, as the personnel dataset does not contain a specific indicator noting whether the principal was newly hired or not.For principals employed in 2002/03, for example, I extracted their personnel records from 2000/01 and 2001/02.I compared their duty root code in each of the previous years to determine whether they had been newly hired.I considered a principal newly hired if their duty root code was not 21 through 24 in both of the previous years.For principals and vice principals who were newly hired, I amended the dataset with a new field that indicated their status as a new administrator.If a principal was employed as an administrator in either of the previous years, I amended the database with a new field that included their status as a current administrator.
The remaining calculations were completed using Microsoft Excel Pivot Tables.I constructed pivot tables to address each of the analytic questions.First, I calculated the number of new principals employed and differentiated whether they were employed as an elementary or secondary principal.I summed the total student enrollment so that a per pupil amount could be calculated.Next, I calculated the number of new principals working in high poverty schools.I identified these principals using the percentage of students in the school who qualified for free-orreduced price lunch.I considered any school with more than 50 percent of the students eligible for free or reduced price to be a high poverty school.

Establishing Costs
Consistent with Knight (2012), this analysis sought to determine the amount coaching support costs.To estimate the cost of the coaching support, I assumed that coaches were paid an annual stipend for each coachee they supported.This arrangement seems common in the literature on state level coaching programs (Villani, 2006).Based on the existing research on state-level coaching programs, I assumed that a coach might receive between $1,625 and $5,000 per coachee (Lochmiller, 2014;Villani, 2006).Other costs, including administrative overhead, would be incurred in addition to this amount.For simplicity, I assumed that the administrative overhead for the coaching program was fixed at $200,000.Within this amount, I included the salary and benefits for a full-time program coordinator ($72,500 salary plus 36% for health insurance and retirement benefits), costs to provide training for the coaches ($25,000), and reimbursements for travel to and from the training sessions and individual coaching sessions ($75,000).I also included approximately $1,400 for supplies and miscellaneous expenses.This estimate was based on previous research about state level coaching programs (Villani, 2006), as actual costs for a statewide coaching program in Washington were not available.

Results
Based on my analysis of state personnel data, I determined that Washington state school districts hired between 394 and 568 new school administrators per year between 2002/03 and 2006/07.Across the five-year period, I found that school districts hired an average of 480 new administrators per year.Elementary school principals were hired in the greatest number, averaging 169 new hires per year compared with 94 new hires for secondary school principals.As illustrated in Table 1, I found that 515 new administrators were hired in 2002/03.Of these, 168 (33 percent) were elementary school principals and 113 (22 percent) were secondary (middle and high school) principals.In 2006/07, 524 new administrators were hired, including 185 (35 percent) elementary school principals and 105 (20 percent) secondary principals.Across each of the years analyzed, a third of the new administrators worked as elementary school principals, while 16 to 22 percent worked as secondary school administrators.Next, I calculated the amount that it would cost to provide coaching support to all of the elementary and secondary school principals.The estimates provided below do not include the program overhead.Instead, the costs illustrated in Tables 2 and 3, represent the cost to compensate coaches to provide coaching support for newly hired elementary and secondary principals.As shown, the cost varies depending on the amount of compensation provided to the leadership coaches, as well as the number of new administrators hired each year.Previous studies of state-level coaching programs indicate that coaches typically receive between $1,625 and $5,000 per coachee (Villani, 2006).As illustrated in Table 2, assuming coaches were paid $5,000 for each coachee they supported, the cost to provide coaching support to elementary school principals would be approximately $845,000 per year or approximately $12.35 per pupil.This amount could be reduced by lowering the stipend to $3,250 per coachee.Under this scenario, the state would spend approximately $549,250 per year or about $8.02 per pupil to provide coaching support to newly hired elementary principals.Finally, reducing the stipend to $1,625 would further reduce the cost to the state.Assuming coaches were paid $1,625 per coachee, the state would spend approximately $274,625 per year or about $4.01 per pupil.Each of the preceding examples assumes that the state covers the full cost of the coaches' compensation.Should districts be required or expected to contribute, the amount would likely vary significantly from one district to the next, as districts hire different numbers of principals each year.As demonstrated in Table 4, districts across the state would spend between $4,390 and $13,509 per year in 2002/03 to provide coaching support to newly hired principals.The largest urban school districts would likely face the most significant costs to provide support, as they typically hire the most new principals each year.Urban districts, which tend to hire more new principals each year, would bear a larger portion of the cost.In Seattle Public Schools, the state's largest school district, the cost to provide coaching support for newly hired principals would range from $50,000 to $145,000 per year if coaches were compensated with a $5,000 stipend.If coaches were paid a smaller stipend, the cost would range from $16,250 to $47,125 per year.Recognizing that states may be unable to bear the full cost of coaching support for all new principals and that districts may be unwilling to contribute local education funds to defray some of the costs, I also calculated the amount to provide coaching support for principals working in highpoverty schools.Previous research indicates that principals in high poverty settings are often the most likely to leave the profession (Beteille, Kalogrides, & Loeb, 2011;Miller, 2009).Thus, providing support only to principals working in the most challenging settings represents a targeted alternative to providing coaching support to all novice principals in the state.My assumption was that a principal working in a high-poverty school would face increased pressures compared to their colleagues working in more affluent settings.As illustrated in Table 5, between 2002/03 and 2006/07, there were between 67 and 110 new administrators working in high poverty schools.

Discussion and Implications
The estimates presented in this article reflect the cost to provide coaching support for novice school administrators in Washington.Based on the estimates, coaching support appears to be financially viable when provided to all newly hired principals, but may be particularly so when targeted to support new administrators working in high-poverty settings.The estimates I derived suggest that it would cost the state as little as $143,975 per year or $4.20 per pupil to provide coaching support to administrators working in high poverty settings.This amount is similar to the amount invested in other state-level programs (Villani, 2006).Providing support to all administrators is also comparable to other state programs.The main advantage to providing coaching support to all administrators is that it ensures that administrators have equitable access across the state to meaningful induction support.Past research indicates that the quality and availability of induction support for principals varies from district to district and school to school (Gray, Fry, Bottoms, & O'Neill, 2007).Statewide programs potentially limit variations in the availability of induction support for novice administrators, yet these programs are often cut when budget reductions loom (Daresh, 2004).
While the development of statewide coaching programs may be viable financially, their development does raise significant new questions and potential policy challenges.First, deploying coaches at-scale would require a robust effort to recruit and train coaches in best practices.Yet, as previously stated, the research base regarding effective coaching practices is limited (Crow, 2012).Despite numerous models and approaches described in the literature (c.f., Bloom, et al., 2005;Costa, Garmston, Anderson, & Glickman, 2002;Knight, 2012), few of the approaches described can be considered 'research-based,' meaning their efficacy has not been subjected to rigorous empirical assessment.The research that does exist suggests that coaching coupled with teacher feedback can yield significant improvements in principal practice, as perceived by teachers who work with the administrators (Bickman, et al., 2012).Thus, the lingering question for policymakers and program leaders is what, specifically, coaches should know and be able to do in order to effectively support principals in becoming effective instructional leaders.
Beyond knowing whether coaching is the most effective professional development strategy and would thus justify substantial public investment, another concern relates to the ability of states to monitor the quality of coaching practices through regulations or professional certifications.While most positions in public education involve some form of professional certification, states have not widely adopted certification criteria for coaches.This is a concern that applies not only to public education but also to the coaching field more generally.According to a recent report of more than 1,200 coaches that was commissioned by the International Coaching Federation (2012) and conducted by Price Waterhouse Coopers, more than 42 percent of respondents in North America indicated that the coaching field required greater regulation.
Finally, as past research has indicated (c.f., Daresh, 2004), state policymakers have been reluctant to target induction support to a single employee group.This reluctance poses significant challenges to both launching a statewide coaching program for administrators, as well as to providing adequate funding to ensure that the program can achieve its stated and/or intended objectives.Fortunately, the research indicates that these challenges can be overcome as programs providing induction support to administrators have been launched in a number of states (Villani, 2006).Moreover, there is ample evidence to suggest that policies are already in place that require districts and states to provide induction support to new administrators (Goldrick, Osta, Barlin, & Burn, 2012).Amending these policies to require that the induction support involve some form of coaching or mentoring may be possible if sufficient evidence suggests that coaching was the most effective form of professional development for novice administrators.While there is qualitative and quantitative evidence suggesting that coaching may be beneficial (Silver, Lochmiller, & Copland, 2009;Lochmiller & Silver, 2010;Bickman, et al., 2012;Grissom & Harrington, 2010), further research is needed before coaching can be conclusively identified as the most effective professional development strategy.
Ultimately, the estimates provided in this article indicate what it would cost to provide coaching support for novice administrators in Washington state.These costs indeed will vary depending on the state context, program model, and coaching fees involved.Despite these variations, the results of the analysis provide hope for those who are seeking to improve support for novice principals by providing coaching to those who are new to their positions and seeking support from an experienced guide.Most importantly, the results of this analysis provide compelling new evidence for policymakers as they consider expanding coaching as a professional development strategy for school administrators.In particular, the results provide per pupil costs, which can inform policy-related discussions.

About the Author
Chad R. Lochmiller Indiana University clochmil@indiana.eduChad Lochmiller is an assistant scientist at the Center for Evaluation and Education Policy at Indiana University, Bloomington.His research examines contemporary education policy issues related to human resource management, resource allocation, and instructional leadership.His research seeks to operationalize different dimensions of leadership practice in schools and districts such that preparation and support for leaders can be enhanced.His interest in leadership coaching relates both to the ways in which coaching support for administrators can be funded, as well as the ways in which coaching can be conceptualized as a unique form of instructional leadership practice.

About the Guest Editor
Sarah Woulfin University of Connecticut Sarah.Woulfin@UConn.eduSarah Woulfin is an assistant professor of Educational Leadership at the University of Connecticut, Storrs.She studies the relationship between education policy, leadership, and instructional reform.Using lenses from organizational sociology, she investigates how leaders influence teachers' responses to reform efforts.In her doctoral work at the University of California, Berkeley, she focused on institutional theory, policy implementation, and coaching.She has published in the American Educational Research Journal (AERJ) and Reading Research Quarterly.Currently, she is an associate editor for Educational Administration Quarterly (EAQ).She is also on the executive steering committee of the Districts in Research and Reform SIG at AERA.From 2009-2012, Dr. Woulfin served as the program chair for AERA's Organizational Theory Special Interest Group.As a former urban public school teacher and reading coach, she was dedicated to strengthening students' reading and writing skills to promote educational equity.As a scholar, her commitment to raising the quality of instruction motivates her research on how policy influences-and is influenced by-administrators and teachers.

Table 1 .
Number of new administrative hires in Washington 2002/03 through 2006/07

Table 2 .
Cost to provide coaching support to newly hired elementary school principalsUsing the same compensation scheme, coaching support for newly hired secondary principals would cost the state $471,000 per year or $8.11 per pupil if coaches were paid $5,000 per coachee.If coaches were paid $3,000 per coachee, the cost would be approximately $306,150 per year or $5.27 per pupil.Finally, if coaches were paid $1,625 per coachee the cost would be approximately $153,075 or $2.63 per pupil.

Table 4 .
Average cost per year to provide coaching support by individual school districts

Table 5 .
Number of administrators hired each year who work in high-poverty schools 2002

/03 2003/04 2004/05 2005/06 2006/07 Average Total new principals hired in Washington
As illustrated in Table6, it would cost the state approximately $443,000 per year or $12.92 per pupil to provide coaching support to elementary and secondary school principals.Reducing the stipend to $3,250 per coachee would reduce the state's cost further to