From Policies to Principals: Tiered Influences on School-Level Coupling

From Policies to Principals: Tiered Influences on School-Level Coupling Abstract For decades, educational scholars have claimed that public schools are loosely coupled organizations, but research does not fully address how schools create the coupling structure. Recent federal policies have centralized the public education system, but state laws, districts, and principals still play a prominent role in shaping the organization of schools. In this paper, I investigate how the multiple tiers of the public education system influence the coupling structures of schools. Utilizing all seven waves of the Schools and Staffing Survey (SASS), I rely on OLS regression to analyze how couplings emerge within schools. Findings suggest three major influences on coupling within schools. First, federal policies have had a non-linear effect on school-level coupling. Second, the relationship between the district and the school affects school-level coupling. Finally, principals play a key role in shaping the coupling within the school, and coupling is a gendered process. Sociologists have described schools as loosely coupled organizations for decades, meaning the subparts of schools are linked and responsive yet largely autonomous (Bidwell 2001; March and Olsen 1976; Orton and Weick 1990; Rowan 1990; Weick 1976). Simultaneously, schools are gendered organizations that largely employ white women in middle-class semi-professions (Acker 1990; Hallett 2007a; Pierce 1995; Williams 1993). Males traditionally assumed leadership roles in schools, but recently the proportion of female principals has risen. This demographic shift makes gendered differences in leadership among principals increasingly salient. To date, these separate threads of research on organizational coupling and gendered leadership in schools have not spoken to each other. As accountability policies and national angst about the quality of our schools becomes a public concern, the work of teachers and their managers (principals) collides with the way schools are organized. It is necessary to integrate these literatures to better understand when couplings change, and this requires an approach that looks at the multiple levels of the public school system. In recent decades, accountability in the US public education system has increased. State and local governments remain important forces for public schools, but large-scale federal changes may drive changes in coupling between principals and teachers. Standards-based reform increased accountability demands on public schools by creating a more centralized federal structure (McDermott 2011). Federal policies have the potential to shift the structure of public schools, but schools are managed by principals, and principals must enact the federal policies in order for the policies to provoke change within schools (Hallett 2007a; Jennings 2010; Spillane et al. 2002; Spillane, Parise, and Sherer 2011). Principals play a key role in negotiating the relationships between policies and teachers (Lee, Smith, and Cioci 1993; Price 2012). As policies enforce more accountability, testing, and oversight, principals can influence how these policies are adopted and navigated by teachers (Diamond 2007). Over this same time period, women started to break the glass ceiling and gain jobs as principals—from 21 percent in the mid-1980s (NCES 1994) to 51 percent in recent years (NCES 2013). The increase in female principals demands we pay attention to the literature on gendered leadership, and opens up new avenues of inquiry about the intersection between different levels of the school system that affect the inner workings of schools. Research on organizations and leadership tells us that men and women will lead differently (Bossert et al. 1982; Fauth 1984; Johnson, Busch, and Slate 2008; Lee, Smith, and Cioci 1993), and there are noteworthy differences between male and female principals’ leadership and management styles (Lee, Smith, and Cioci 1993; Price 2012). It is plausible that these female principals will enact coupling structures at the school level differently from their male colleagues. How do the multiple levels of the public school system affect school-level coupling? And in what context—federal, state, and/or local—does gender matter for coupling? Coupling is a central issue to the sociology of education, and I revisit foundational questions from a new conceptual and methodological approach. I combine two major literatures that address how schools are organized—coupling and gendered leadership. This is the first large-scale quantitative study to examine coupling from the federal level to the schoolhouse door. My analyses of nationally representative data provide a more nuanced picture of what coupling looks like in public schools across two and a half decades and two major policy eras. Coupling in Schools Coupling refers to how closely formal organizational structures (e.g., policy) are related to the real technical activities occurring within the organization (Weick 1976). Schools are loosely coupled when principals and teachers are tied to one another, but retain a relatively great degree of independence. For instance, when teachers have autonomy and control over their classrooms, and are not tightly connected to their principals, the schools are more loosely coupled. But if principals exert a great deal of influence over teachers’ classrooms, then the schools are more tightly coupled. Coupling is not a binary, however, where schools are either loosely coupled or tightly coupled. Rather, scholars have highlighted the dialectical nature of leadership practice within schools, which implies that coupling is not unidimensional. This understanding allows us to focus on how some dimensions are more tightly coupled, while others are more loosely coupled. In fact, work has suggested that coupling on some dimensions, such as instructional content (especially in math and language arts), is tighter than on other dimensions, such as pedagogical decisions (Diamond 2007). Weick (1976) labels schools as loosely coupled organizations and explains how loose coupling protects individuals and subunits of an organization during times of uncertainty. It may not pervade the entire organization or affect the organization at all times, however, and some areas of the organization may have tighter couplings in particular eras while other areas of the organization have looser couplings during other policy eras. Neo-institutionalists argue that schools attempt to maintain legitimacy by engaging in ceremonial displays of compliance, but do not alter the actual activities in classrooms (Meyer and Rowan 1977, 1978). These conceptualizations help organizational scholars make sense of the fact that schools’ structures remained largely unchanged despite large-scale policy changes. A review of the literature differentiates between eight types of loose coupling: “loose coupling (1) among individuals, (2) among subunits, (3) among organizations, (4) between hierarchical levels, (5) between organizations and environments, (6) among ideas, (7) between activities, and (8) between intentions and actions” (Orton and Weick 1990, 208, numbers added). Here, I focus on loose coupling in schools’ hierarchical levels (e.g., between the principal and teachers in schools) (Firestone 1985; Gamoran and Dreeben 1986). When educational scholars assume that schools are loosely coupled, they are often referring to the relationship between principals and teachers, but other aspects of a school may be tightly coupled even when principals and teachers are loosely coupled. Weick (1976) also points out that loose coupling in one arena of an organization suggests tight coupling elsewhere. Previous work on hierarchical levels does not fully capture all the levels of our public school system. I build upon this literature, focusing on hierarchical coupling at three distinct levels. First, I define micro-level coupling as the formal relationship between the principal and the teachers within the school. Second, I define meso-level coupling as the formal relationship between the school district and the school. Finally, I consider macro-level coupling as the relationship between accountability policies and the school. Macro structures include federal and state policies within this level of the tiered school system. My conceptualization of coupling allows me to consider the various levels of the school system and the roles they play in shaping micro-level coupling. By taking multiple levels into account, we begin to understand how some levels may tighten coupling, while other levels loosen coupling, even within the same time period. Since the 1970s, studies have suggested that schools are loosely coupled and have been throughout history (Bidwell 2001; Gamoran and Dreeben 1986; March and Olsen 1976; Orton and Weick 1990; Rowan 1990; Weick 1976). In general, scholars underscore the principal-to-teacher relationship (i.e., micro-coupling), and indicate that principals exclude themselves from many day-to-day teaching and learning activities. Teachers oversee the lion’s share of instructional decisions (1978). Teachers do not have much control over school-level decisions (Lee, Dedrick, and Smith 1991; Renzulli, Parrott, and Beattie 2011), but they do strongly desire and seek to preserve autonomy over classroom affairs (Hanson 1977; Lortie 2002). The emphasis on autonomy, control, and input over classroom behavior signifies the sizable importance of relationships between teachers and principals in the loose coupling discourse. Over the past twenty years, various studies have suggested a surprising trend from loose to tight coupling (recoupling) within schools. Specifically, teachers report a decline in autonomy over classroom and school-related decisions (Coburn 2004; Diamond 2007, 2012; Hallett 2010; Ingersoll 2003; Young 2006). For example, teachers report “turmoil” when their administration replaces a loosely coupled system with tight coupling in response to policy changes (Hallett 2010). Teachers favor loosely coupled school environments when faced with increased institutional pressures (Coburn 2004; Ingersoll 2003), but studies have not fully investigated how the combined role of macro- (federal and state) and meso-level (local) institutional structures are catalysts for tighter or looser coupling on a national scale. I contribute to the discussion of coupling in organizations by using a tiered approach to understand how schools are coupled over several policy eras. Empirical studies of coupling have focused on the consequences of loose or tight coupling and have explored their sources. Research underscores why loose coupling is ideal for organizations (Cohen and March 1974; March and Olsen 1976; Meyer and Rowan 1977; Rowan 1981; Sauder and Espeland 2009), why teachers prefer loose coupling (Ingersoll 2003; Lortie 2002), and how teachers attempt to regain a loosely coupled structure when faced with threats to loose coupling (Coburn 2004). The impetus behind organizational couplings and the general trends of couplings at a national level is necessary, and the more recent findings on tighter coupling and recoupling motivate this study of how various tiers of the school system affect coupling. In order to fully understand how various tiers of the public school system affect coupling within public schools, we need to attend to accountability structures and district-school relationships. As noted above, we also need to engage the literature on gendered leadership. I use multiple levels of the educational system—federal, state, local, and principal—to offer a comprehensive analysis of gendered leadership and coupling in public schools. Accountability through Policy Multiple levels of the public education system are charged with producing and maintaining coupling structures, with much attention focused on the role of high-stakes accountability reforms handed down from various levels of government in recent years (Coburn 2004; Desimone 2013; Diamond 2007; Elmore, Abelman, and Furman 1996; Hallett 2010; Spillane and Burch 2006). In 1983, A Nation at Risk spurred the standards-based reform movement with claims of a “‘rising tide in mediocrity’ in U.S. Schools” (McDermott 2011, 60). The Clinton administration subsequently brought about the reauthorization of the Elementary Secondary Education Act (ESEA) as the Improving America’s Schools Act (IASA), and the Bush administration reauthorized ESEA/IASA as the No Child Left Behind Act of 2001 (NCLB). Unlike their predecessors, these policies redirected attention toward equity in education vis-à-vis performance accountability, and away from issues of desegregation, access, and equality in spending. Such federal policies and reports are fundamentally designed to affect all public schools. Geared toward performance accountability and improving academic achievement in public schools, policymakers assume that policy efforts will result in tangible changes or outcomes at the school level. However, many broad policies fail to take form within the walls of schools (Eagly and Johnson 1990; Pitner 1981; Shakeshaft 1987). While the federal policies from the mid-1980s through the first decade of the twenty-first century had variation in objectives, the overarching reforms centered on an increase in academic accountability standards for public schools across the nation, and are thus referred to as the standards-based reform movement (McDermott 2011). The federal policy reforms put forth by IASA and NCLB are highly focused on closing the achievement gap for racial minorities and poor students (Grissom, Kalogrides, and Loeb 2013), suggesting that the largest impact will be felt among schools that serve a greater proportion of minority and lower-class students. Organizational scholars recognize a changing educational landscape, and qualitative studies document a recoupling (Coburn 2004; Davies, Quirke, and Aurini 2006; Hallett 2010). Standards-based reforms have been specifically credited for changing how teachers deal with struggling students, teaching to the test, and classroom content (Desimone 2013). With few exceptions, contemporary educational research attributes the shifting trend in coupling structures (from looser to tighter), at least in part, to macro-level structures (e.g., federal policy, state accountability structures). Given the explicitly stated goals of federal policies, to increase academic accountability among public schools, I hypothesize that federal-level policies will tighten coupling between principals and teachers. State-level policies are also influential in shaping schools. State-level accountability structures can vary widely depending on the political climate of the state. Federal policies request test scores (e.g., NCLB’s Adequate Yearly Progress—AYP), but states have the power to increase standards, test more frequently, or standardize curriculums. States differ broadly along a continuum on which some states are low in high-stakes testing, others exhibit a great deal of high-stakes testing and accountability to the state, and still others fall somewhere in between (Carnoy and Loeb 2002). States retain a great deal of control over public schools and can dictate standards above and beyond the federal-level mandates. Though not as pervasive as federal-level policies, state accountability structures and laws often diffuse across states (for an example of charter school laws, see Renzulli and Roscigno [2005]). State governments will adopt similar policies (i.e., to their neighboring states), resulting in homogeneous regulations. While the intent of external pressures is to yield increased results in academic performance, it is possible that the degree of external pressures results in tighter coupling within the school, as teachers and administrators negotiate the requirements of the state. I hypothesize that state macro-level accountability structures play a role in tightening the micro-level coupling structures within schools. The Local Context Local governments and school boards oversee many organizational aspects within schools and still wield substantial control over their local schools (Diamond 2007), despite concerted centralization efforts by the federal and state governments. Macro structures at the federal and state level should trickle down to local schools, but local school boards and governments can require schools to adhere to additional district-level guidelines. For example, local governments may influence hiring decisions, curriculum content, or discipline policies. In these circumstances, the local district may play a central role in determining the internal structure of the school, and can serve as an intermediary between macro-level policies and changes within the school. Thus, various levels of the school system simultaneously affect day-to-day activities within schools, although they may affect different aspects of the school system. Local school districts often attempt to play an active role in shaping the internal structure of schools, and can penetrate the core activities inside classrooms (Coburn 2004; Hallett 2010). Such interplay between various hierarchical levels of the school system lend further credence to Orton and Weick’s (1990) proposition that we understand coupling in organizations through a dialectical approach. For instance, Hallett (2007b) found that the local school district will attempt to influence the organizational structure of a school by hiring a principal whose leadership style most closely aligns with their goals. Thus, the local tier of the school system may apply indirect pressures when placing administrators in schools. Neo-institutionalists would expect principals and teachers to decouple when faced with pressure from the local school district, and while some teachers do turn to decoupling strategies when faced with these institutional pressures, many others assimilate (among other strategies) (Coburn 2004). Moreover, the principal may work closely with the local district when the district is more integrated in the decision-making that affects the school. The local government is the meso-level of coupling, connecting the macro-level (federal and state) structures and the micro-level (school). While neo-institutionalists would predict that external pressure from the school district would result in looser coupling in schools, more recent studies on coupling suggest that district pressures may actually tighten coupling in the school (Diamond 2012). This may be especially true if principals actively push/support the school district’s agenda. Given the trends in recent literature, I hypothesize that tighter coupling from districts to schools will tighten the coupling between principals and teachers. The Role of the Principal: Female Leaders The institutional environment helps shape coupling in schools, but the principal plays a pivotal role in interpreting and enacting the mandates (Coburn 2004; Diamond 2007, 2012; Hallett 2010; Seyfarth and Bost 1982; Spillane et al. 2002; Young 2006). Principals are responsible for mediating policy messages, and this may disrupt teachers’ preferred structure within the school (Weiss and Cambone 1994). Even when the macro- or meso-level policies are touted as the overarching factor in shifting organizational structure of schools, the qualitative data illuminates the role of the principal (Diamond 2012; Spillane et al. 2002; Spillane, Parise, and Sherer 2011). Teachers commonly report losing autonomy (Diamond 2007; Hallett 2010) because principals do not let them “do their jobs” (Hallett 2010, 58). According to teachers, principals endorse and enforce policies within the school, exemplifying Orton and Weick’s (1990) dialectical understanding of coupling. While teachers report that principals do influence their classroom decisions, they also highlight how other forces are influential factors, such as other teachers, standards, testing, textbooks, and the Internet (Diamond 2007). Thus, principals do possess some agency to enact policy standards and changes, but if male and female principals embody different leadership styles, then recoupling or decoupling could also be a result of gendered leadership. A focus on the inner workings of schools—particularly principals—cannot ignore gender. Schools are gendered organizations; the gendered norms of schools disproportionately place males in administrator positions and females in teaching positions (Hallett 2007a; Lee, Smith, and Cioci 1993). Women have become principals in increasing numbers, making gendered leadership an increasingly critical issue in public schools. Principals display some of the same gendered management styles found in other organizational settings (Eagly and Karau 2002; Weiss and Cambone 1994). Female principals are more likely to communicate with teachers, stop into classrooms, walk through the hallways, and know the general pulse of the school (Fauth 1984; Ingersoll 1996; Pfeffer 1981; Shakeshaft 1987). Female principals engage in these behaviors due to their general interest in the teachers in their school. In contrast, males engage in traditional non-participatory management styles, relying on authoritative directives that are not followed up with communication, trips around the school, or time spent in classrooms (Lee, Smith, and Cioci 1993). In general, female principals wish to be “more prominently and personally involved,” compared to their male counterparts (Charters and Jovick 1981, 322). And while a democratic leadership style is often considered more effective (Eagly and Johnson 1990), the management of schools has shifted from a focus on educational philosophies toward that of bureaucracy and efficiency, an approach more consistent with other types of organizations, and may favor male principals. Based on accounts of leadership styles, I expect coupling in schools headed by female principals to be tighter than in those headed by males.1 Principals are in positions of power, and must acquire legitimacy for their leadership role while simultaneously maintaining organizational legitimacy for the school (Spillane, Hallett, and Diamond 2003). In some cases, this may only require assuming the title of “principal,” but this is unusual and principals tend to garner legitimacy when they can exhibit valued interaction styles (Spillane, Hallett, and Diamond 2003). When seeking legitimacy, female principals could turn to policies and regulations already viewed as legitimate in order to gain control over a school. Organizational actors can acquire legitimacy through sources of authority (Fauth 1984; Tyler 2006; Wingfield 2009; Zelditch 2001); for principals, authority may come from higher powers (e.g., policies, local government), and a principal could choose to embrace and implement policies in order to glean legitimacy (Spillane et al. 2002). Principals may also wish to send a signal to those individuals who oversee their work (district-level administrators). Districts want to hold schools accountable, but require a principal who is willing to enforce policies and maintain surveillance over the school (Hallett 2007b, 2010). Female principals may feel pressure to actively implement district policies in order to acquire legitimacy from those at the district level, but Hallett (2007b, 2010) demonstrates how principals who advocate for more tightly coupled relationships with teachers may face opposition and actually lose legitimacy with teachers. Context and circumstances may contribute to differences in gendered leadership because gendered leadership does not exist outside sociopolitical and local context. Multiple hierarchical levels can function harmoniously to shape the organizational coupling structure of schools; thus, I explore how female principals affect micro-coupling during federal policy eras with high accountability and in districts where tighter coupling exists between the district and the school. Data I draw on several data sources; primarily, the Schools and Staffing Survey (SASS), from the National Center for Education Statistics (NCES). This is the best dataset for a study of school trends in organizational structure and activities because it is a large, nationally representative sample of schools, principals, and teachers with measures of institutional environments and organizational structure. SASS includes seven waves of data spanning two and a half decades and multiple socio-political eras. The analyses shown in this manuscript draw on all seven waves of the SASS data (1987–1988, 1990–1991, 1993–1994, 1999–2000, 2003–2004, 2007–2008, 2011–2012). I refer to waves using the first part of the academic year (e.g., 1987–1988 is 1987). Due to changes in the questionnaire for the most recent wave of data (2011–2012), I am unable to include this wave in all of my analyses. I use this wave of data whenever possible, and in an appendix I include separate analyses that utilize this wave of data, but there are missing variables. When discussing findings, I also comment on how the last wave affects the results. The SASS data is a repeated cross-section of randomly sampled schools, resulting in a nationally representative sample of schools in each wave. For theoretical and empirical reasons, I exclude some schools and teachers from my analyses. After excluding all schools that are not traditional public schools (i.e., private schools, alternative schools, charter schools, or non-traditional schools), and all teachers in non-traditional classrooms, I am left with approximately 34,940 school-years (25,640 unique schools, with 7,040 repeatedly sampled) across six waves of data.2 For analyses that include the 2011 wave of data, I have approximately 40,360 school-years (28,910 unique schools, with 8,740 repeatedly sampled). Given the large number of cases, I also ran models using a random 10 percent subsample. Results of these models mirror those presented here. In a unique combination of data sources, I link the SASS with supplementary socio-political data that allow me to consider environmental factors that could influence schools. By including variables that address the external environment, I am better able to understand how the socio-political environment plays a role in shaping the organizational structuring of schools. Variables Dependent Variable Micro-Level Coupling (Loose-to-Tight Coupling) Micro-level coupling is a continuous variable, ranging from 0 to 4, where lower numbers indicate looser coupling and higher numbers indicate tighter coupling. Operationalizing a concept that is traditionally found in qualitative literature is a difficult task; thus, I focus on the elements highlighted in the scholarship on school coupling to create this variable. The primary signs of coupling can be found in elements of control and autonomy (i.e., how tightly linked administrative control is over what teachers do in their classrooms). This is consistent with previous research that studies the opposing concepts of connection and autonomy (Orton and Weick 1990), and recent educational research on coupling (Hallett 2010). This school-level coupling index is composed of five items indicating the extent of control/influence teachers have over: (1) choosing textbooks and other instructional materials, (2) choosing content, topics, and skills to be taught, (3) selecting teaching techniques, (4) disciplining students, and (5) determining the amount of homework assigned. These five items largely capture academic instruction, and are likely the places where change would occur if schools respond to the federal goals of performance accountability. Items are reported by teachers within each school in each of the waves. Teachers report how much individual influence/control they possess, and teachers responded to a Likert scale that ranged from “complete control/a great deal of control” (loosely coupled) to “no control” (tightly coupled). The SASS questionnaires asked identical questions across waves of data but offered different ordinal response categories to respondents in some waves of data. I standardized each item to have a 0–4 scale by multiplying it by the appropriate scaling factor (see table 1 for scaling factors) and averaged items to form a continuous micro-coupling index (α = 0.73). Schools are not just tightly or loosely coupled; they can be coupled on some dimensions and not others even in the same policy era. Therefore, I consider these five items as separate dependent variables. I do not show the models, but all models are substantively similar, ensuring that one or two items from the index are not driving the results presented here. Table 1. Scaling Factors for Standardizing Scales Original  Scaling factor  Rescaled  Waves: 1987, 1990, 1993  6-point scale  6 to 5  5-point  Range: 0–5    Range: 0–4  0  4/5  0  1  4/5  0.8  2  4/5  1.6  3  4/5  2.4  4  4/5  3.2  5  4/5  4  Original  Scaling factor  Rescaled  Waves: 2003, 2007, 2011  4 point Scale  4 to 5  5 point  Range: 0–3    Range: 0–4  0  1 1/3  0  1  1 1/3  1.34  2  1 1/3  2.67  3  1 1/3  4  Original  Scaling factor  Rescaled  Waves: 1987, 1990, 1993  6-point scale  6 to 5  5-point  Range: 0–5    Range: 0–4  0  4/5  0  1  4/5  0.8  2  4/5  1.6  3  4/5  2.4  4  4/5  3.2  5  4/5  4  Original  Scaling factor  Rescaled  Waves: 2003, 2007, 2011  4 point Scale  4 to 5  5 point  Range: 0–3    Range: 0–4  0  1 1/3  0  1  1 1/3  1.34  2  1 1/3  2.67  3  1 1/3  4  Note: All waves are scaled to the 1999 wave of SASS data. Teachers primarily consider their own classrooms, and the intra-class correlation among teachers (i.e., teachers who teach in the same school) is significant in all waves, ranging from 11 to 17 percent. This indicates the degree to which teachers agree on the organization of activities and autonomy within their school. To obtain a school-level measure of coupling, teacher reports of coupling are aggregated to the school level by averaging teachers’ reports within each school. On average, four to five teachers were interviewed from each school. The variation in teachers’ responses within the same school is less than the variation in teachers’ responses across schools; therefore, I am confident in the validity of using the aggregated responses of teachers to form a school-level variable. Federal-Level Independent Variables Policy Eras Time is an important dimension of this study, and I use the survey year as an indicator of policy era. I enter the variable as categorical to assess non-linearity in its effects, although for initial analyses on the trends of micro-coupling, I use a continuous policy-era variable that ranges from 0 to 6. Federal-level policies have changed over time. Survey year 1987 is the first wave of data available after the 1983 report A Nation at Risk, and the year 1987 is a reference category in all analyses. Survey year 1999 is the first wave of data after ESEA was reauthorized as IASA. Survey year 2003 is the first wave of data in the NCLB era, and while survey year 2007 still falls under the auspices of NCLB, much like 1990 and 1993 are still directly following A Nation at Risk, it is six years after the controversial policy, and schools are no longer transitioning into NCLB requirements. For those analyses that include the final wave of data, year 2011 is well after the enactment of NCLB, and by this point many states have sought waivers in place of NCLB requirements (US Department of Education 2013). SASS data are ideally situated for an analysis that considers the relationship between policy eras and micro-coupling. I expect time lags between policy implementation and micro-coupling, and if the waves of data are concurrent with policy changes, then the analyses may not truly capture the effects of the federal policy eras. NCLB passed in 2001, and was signed into law in early 2002; therefore, the 2003–2004 wave of SASS data is useful for capturing the early transitions into NCLB while still allowing for a lag between policy implementation and changes to micro-coupling. State-Level Independent Variables Index of High-Stakes Testing I am interested in accountability structures that states impose to improve test scores or control schools from the macro level. For accountability at the state level, I use Carnoy and Loeb’s (2002) state-accountability index. This index signifies the degree to which a state participates in high- or low-stakes testing. It “captures the degree of state external pressures on schools to improve student achievement according to state-defined performance criteria” (Carnoy and Loeb 2002, 311). This index ranges from 0 to 5, with 0 indicating a very low level of accountability within the state and 5 indicating a great deal of accountability. Though not time-varying, this is the best indicator available for accountability at the state level. Election Results by State I use the election maps provided by the US Electoral Maps (Office of the Federal Register 2013) to gauge the relative liberal or conservative leanings of each state. I coded the direction each state voted in the presidential election prior to the wave of data presented. (e.g., models with the 2007 wave of data use the 2004 presidential election). Each state is coded as 1 or 0, with 1 indicating that the state voted for the Republican candidate. (This is consistent with research that links state characteristics with educational data. See Renzulli and Roscigno [2005].) State Charter School Laws I use a dichotomous variable to indicate whether or not the state has a law permitting charter schools. Charter schools are not included in my analyses, but this information helps signal how states view educational choice movements. States adopted charter school laws in different years, and some never adopt charter school laws, making this a time-varying variable. Local-Level Independent Variables Meso-Level Coupling To capture meso-level coupling—coupling between the school district/school board and the school—I create an index consisting of three variables indicating the district’s control over hiring teachers in the school, setting curricula, and setting discipline policies for schools (α = 0.69). I recode this variable using the same format from my dependent variable. The categories range from “no influence” to “major influence,” and the coupling index between the local government and school is a loose-to-tight index, parallel to the micro-level (principal-to-teacher) coupling index (i.e., 0–4), but preliminary examination of this variable revealed non-normality. Preliminary models demonstrate non-linearity in the association between meso-level coupling and micro-coupling. Given this non-normality and evidence of non-linearity, I divide the meso-level coupling variable into quartiles, and enter this variable into the analyses as a categorical predictor. Quartile comparisons are then made via Wald tests (using alternative out-groups/reference categories). These tests reveal that the bottom three quartiles do not differ from one another in their association with micro-level coupling, but they each differ significantly from the fourth quartile in their associations. Hence, I account for this non-linearity by dividing my meso-level coupling variable into tight coupling (fourth quartile) or not (bottom three quartiles). Unfortunately, these measures are excluded in the most recent wave of data collection (2011), and models using this index are limited to data from waves 1987–2007. Merit Pay/Bonus I use a measure of merit pay to indicate whether or not the local district or school uses a system of performance-based merit pay or bonus to reward teachers for students’ test scores. This is a dichotomous variable (1 = merit pay/bonuses) and refers to the availability of a school-wide bonus for all teachers in a school with exceptional performance/improvement. Principal Characteristics Female Principal’s sex is a dichotomous variable (female = 1). Non-White Principal’s race is indicated with a dummy variable (nonwhite = 1) (White: 86 percent; Black: 7.8 percent; Hispanic: 3.3 percent; Asian: 1.3 percent; American Indian: 1.3 percent). Highest Degree Earned Most principals earn advanced degrees. I include the measure for doctorate degree in my models for highest degree earned, and all other degrees as the reference category. Principal Tenure The number of years a principal served ranges from 0 to 47. Principals who report 0 are serving their first year as principal, and the average is just over 5 years. Teaching Experience More than 99 percent of principals taught prior to becoming a principal. The length of time a principal spent teaching prior to becoming a principal ranges from 0 to 42 years. The average is 11.49 years, although females have a statistically significant higher average than males (13.59 vs. 10.49). Control Variables School-level and environmental factors could play a role in how school employees organize the internal activities of a school. The demographics of a school may motivate administrators, teachers, and staff to tighten or loosen coupling based on needs that are related or unrelated to policies, accountability structures, rewards, or the local school board. Controlling for these school-level factors reduces the likelihood that I am mis-specifying effects actually attributable to other organizational factors. I control for environmental characteristics, such as the location (i.e., northeast, midwest, south, and west), the setting (i.e., urban, suburban, and rural), grades served (i.e., elementary or secondary), enrollment, school demographics (i.e., percent free lunch, percent White, Black, Hispanic, Asian, and American Indian), and the percent of teachers who teach in tested subjects.3 Finally, due to the fact that some schools appear in multiple waves of the data, I also control for schools that are represented more than once over time. Table 2 provides additional information about each variable. Table 2. Variables of Interest: Definitions, Sources, and Descriptives (Means and Standard Deviations for All Seven Waves are Presented in Parentheses)  Variable  Description and Coding  Mean  SD  Dependent Variable  Micro-Level Coupling (Individual items are presented below)  Micro-level coupling is made up of 5 items that indicate the formal relationship between the principal and teachers. The following items are included in my scale: the extent of control/influence teachers have over (1) choosing textbooks and other instructional materials, (2) choosing content, topics, and skills to be taught, (3) selecting teaching techniques, (4) disciplining students, and (5) determining the amount of homework assigned. These items are reported by teachers within each school. Questions ask teachers to report how much individual influence or control they possess and teachers responded to a likert scale that ranged from “complete control/a great deal of control” (loosely coupled) to “no control” (tightly coupled). The scaled variable ranges from 0-4. α = 0.73.  0.91 (0.90)  0.47 (0.47)   1. Textbooks  Selecting textbooks and other instructional materials. Range:0-4. Higher numbers indicate tighter coupling.  1.42 (1.44)  0.87 (0.89)   2. Content  Selecting content, topics, and skills to be taught. Range:0-4. Higher numbers indicate tighter coupling.  1.34 (1.37)  0.85 (0.87)   3. Techniques  Selecting teaching techniques. Range:0-4. Higher numbers indicate tighter coupling.  0.50 (0.51)  0.50 (0.51)   4. Discipline  Disciplining students. Range:0-4. Higher numbers indicate tighter coupling.  0.84 (0.83)  0.58 (0.58)   5. Homework  Determining the amount of homework to be assigned. Range:0-4. Higher numbers indicate tighter coupling.  0.43 (0.43)  0.50 (0.50)  Independent Variables  Federal Level      Survey Waves  Continuous variable ranges from 0 to 6, and represents each wave in the SASS data (1987, 1990, 1993, 1999, 2003, 2007, and 2011).  2.48 (2.94)  1.65 (1.94)  Year 1987  1987-1988 survey from the SASS questionnaires. (A Nation at Riskera) (Reference category).  0.15 (0.13)  0.36 (0.33)  Year 1990  1990-1991 survey from the SASS questionnaires.  0.18 (0.16)  0.39 (0.37)  Year 1993  1993-1994 survey from the SASS questionnaire (IASA era).  0.18 (0.15)  0.38 (0.36)  Year 1999  1999-2000 survey from the SASS questionnaire.  0.17 (0.15)  0.38 (0.36)  Year 2003  2003-2004 survey from the SASS questionnaire (NCLB era).  0.17 (0.13)  0.38 (0.36)  Year 2007  2007-2008 survey from the SASS questionnaire.  0.15 (0.13)  0.35 (0.33)  Year 2011  2011-2012 survey from the SASS questionnaire.  ---- (0.13)  ---- (0.34)    State Level      Index of High Stakes Testing  This index ranges from 0-5, and represents the extent each state has external accountability. 0 indicates a low level of accountability, and 5 is the highest level of accountability. Source: (Carnoy and Loeb 2002).  2.25 (2.26)  1.49 (1.49)  Election Results  Each state is coded as 1 for republican or 0 for democrat based on how they voted in the most recent presidential election. Data from the 2007 wave uses election results from the 2004 election.  0.64 (0.61)  0.48 (0.49)  State Charter School Law  This is a dichotomous variable that indicates whether or not a state has a law that allows charter schools. States adopt charter school laws at different times; thus, this variable changes over time. States with charter school laws are coded as 1 once they acquire the law.  0.41 (0.47)  0.49 (0.50)    District Level      Meso-Level Coupling  This level of coupling is the relationship between the district and the school. This variable is a scale made up of 3 items that range from 0-4, but due to non-normality, I include a categorical variable that indicates if the meso-level coupling is in the highest quartile (very tight) and is coded as 1. α = 0.61.  0.21 (---)  0.40 (---)  Bonus  This is a measure of merit pay that indicates whether or not the local district or school uses a system of performance based merit pay or a bonus to reward teachers for their students’ test scores. This is a dichotomous variable and refers to the availability of a school wide bonus for all teachers in a school with exceptional performance or improvement. The measure of school wide bonuses also considers incentives outside of the school that are present in the district.  0.09 (0.09)  0.29 (0.28)    Principal Level      Gender  Female principals are coded as 1.  0.31 (0.32)  0.46 (0.47)  Race  Nonwhite principals are coded as 1. White principals are coded as 0. Non-white principals only make up 13% of the data.  0.13 (0.13)  0.34 (0.34)  Highest Degree Earned  The highest degree the principal has earned. Degrees range from Bachelors to Masters, Specialist, and Doctorates. I code doctorate degrees as 1, and use all other degrees as the reference category.  0.09 (0.09)  0.29 (0.29)  Number of Years as Principal  The number of years a principal has served as a principal.  5.13 (5.04)  5.40 (5.30)  Number of Years Teaching Prior to Principal  The number of years a principal spent teaching in the classroom prior to becoming a principal.  11.45 (11.49)  6.36 (6.37)  Control Variables  Northeast  School is located in the northeast.  0.16 (0.16)  0.37 (0.37)  Midwest  School is located in the midwest.  0.26 (0.26)  0.44 (0.44)  West  School is located in the west.  0.25 (0.25)  0.43 (0.43)  South  School is located in the south (Reference category).  0.34 (0.33)  0.47 (0.47)  Urban  School is located in an urban setting.  0.20 (0.20)  0.40 (0.40)  Suburban  School is located in a suburban setting.  0.31 (0.31)  0.46 (0.46)  Rural  School is located in a rural setting (Reference category).  0.49 (0.50)  0.50 (0.50)  Grades Served  Grades served in the school. Elementary schools are coded as 1. Combined and secondary schools are coded as 0.  0.55 (0.55)  0.50 (0.50)  Enrollment  The number of students enrolled in the school.  633.67 (642.86)  497.18 (504.32)  Percent Free Lunch  This variable is the percentage of students who are eligible to receive free lunch in the school (range: 0-100).  36.07 (37.16)  26.10 (26.41)  Percent White  Percent of White students enrolled in the school. (Reference category).  73.17 (72.47)  29.99 (29.99)  Percent Black  Percent of Black students enrolled in the school.  12.05 (11.91)  22.07 (21.73)  Percent Hispanic  Percent of Hispanic students enrolled in the school.  8.35 (9.05)  17.46 (18.05)  Percent Asian  Percent of Asian students enrolled in the school.  2.78 (2.79)  9.30 (8.93)  Percent American Indian  Percent of American Indian students enrolled in the school.  3.65 (3.52)  13.54 (13.23)  Teachers in a Tested Subject  The percent of teachers who responded within the school, and teach in a tested subject.  48.54 (49.04)  33.33 (33.17)  Duplicate School  Some schools apper more than once across all waves of the SASS data. I control for schools that appear more than once (coded as 1).  0.49 (0.49)  0.50 (0.50)  N = 34,940 (N=40,360)  (Means and Standard Deviations for All Seven Waves are Presented in Parentheses)  Variable  Description and Coding  Mean  SD  Dependent Variable  Micro-Level Coupling (Individual items are presented below)  Micro-level coupling is made up of 5 items that indicate the formal relationship between the principal and teachers. The following items are included in my scale: the extent of control/influence teachers have over (1) choosing textbooks and other instructional materials, (2) choosing content, topics, and skills to be taught, (3) selecting teaching techniques, (4) disciplining students, and (5) determining the amount of homework assigned. These items are reported by teachers within each school. Questions ask teachers to report how much individual influence or control they possess and teachers responded to a likert scale that ranged from “complete control/a great deal of control” (loosely coupled) to “no control” (tightly coupled). The scaled variable ranges from 0-4. α = 0.73.  0.91 (0.90)  0.47 (0.47)   1. Textbooks  Selecting textbooks and other instructional materials. Range:0-4. Higher numbers indicate tighter coupling.  1.42 (1.44)  0.87 (0.89)   2. Content  Selecting content, topics, and skills to be taught. Range:0-4. Higher numbers indicate tighter coupling.  1.34 (1.37)  0.85 (0.87)   3. Techniques  Selecting teaching techniques. Range:0-4. Higher numbers indicate tighter coupling.  0.50 (0.51)  0.50 (0.51)   4. Discipline  Disciplining students. Range:0-4. Higher numbers indicate tighter coupling.  0.84 (0.83)  0.58 (0.58)   5. Homework  Determining the amount of homework to be assigned. Range:0-4. Higher numbers indicate tighter coupling.  0.43 (0.43)  0.50 (0.50)  Independent Variables  Federal Level      Survey Waves  Continuous variable ranges from 0 to 6, and represents each wave in the SASS data (1987, 1990, 1993, 1999, 2003, 2007, and 2011).  2.48 (2.94)  1.65 (1.94)  Year 1987  1987-1988 survey from the SASS questionnaires. (A Nation at Riskera) (Reference category).  0.15 (0.13)  0.36 (0.33)  Year 1990  1990-1991 survey from the SASS questionnaires.  0.18 (0.16)  0.39 (0.37)  Year 1993  1993-1994 survey from the SASS questionnaire (IASA era).  0.18 (0.15)  0.38 (0.36)  Year 1999  1999-2000 survey from the SASS questionnaire.  0.17 (0.15)  0.38 (0.36)  Year 2003  2003-2004 survey from the SASS questionnaire (NCLB era).  0.17 (0.13)  0.38 (0.36)  Year 2007  2007-2008 survey from the SASS questionnaire.  0.15 (0.13)  0.35 (0.33)  Year 2011  2011-2012 survey from the SASS questionnaire.  ---- (0.13)  ---- (0.34)    State Level      Index of High Stakes Testing  This index ranges from 0-5, and represents the extent each state has external accountability. 0 indicates a low level of accountability, and 5 is the highest level of accountability. Source: (Carnoy and Loeb 2002).  2.25 (2.26)  1.49 (1.49)  Election Results  Each state is coded as 1 for republican or 0 for democrat based on how they voted in the most recent presidential election. Data from the 2007 wave uses election results from the 2004 election.  0.64 (0.61)  0.48 (0.49)  State Charter School Law  This is a dichotomous variable that indicates whether or not a state has a law that allows charter schools. States adopt charter school laws at different times; thus, this variable changes over time. States with charter school laws are coded as 1 once they acquire the law.  0.41 (0.47)  0.49 (0.50)    District Level      Meso-Level Coupling  This level of coupling is the relationship between the district and the school. This variable is a scale made up of 3 items that range from 0-4, but due to non-normality, I include a categorical variable that indicates if the meso-level coupling is in the highest quartile (very tight) and is coded as 1. α = 0.61.  0.21 (---)  0.40 (---)  Bonus  This is a measure of merit pay that indicates whether or not the local district or school uses a system of performance based merit pay or a bonus to reward teachers for their students’ test scores. This is a dichotomous variable and refers to the availability of a school wide bonus for all teachers in a school with exceptional performance or improvement. The measure of school wide bonuses also considers incentives outside of the school that are present in the district.  0.09 (0.09)  0.29 (0.28)    Principal Level      Gender  Female principals are coded as 1.  0.31 (0.32)  0.46 (0.47)  Race  Nonwhite principals are coded as 1. White principals are coded as 0. Non-white principals only make up 13% of the data.  0.13 (0.13)  0.34 (0.34)  Highest Degree Earned  The highest degree the principal has earned. Degrees range from Bachelors to Masters, Specialist, and Doctorates. I code doctorate degrees as 1, and use all other degrees as the reference category.  0.09 (0.09)  0.29 (0.29)  Number of Years as Principal  The number of years a principal has served as a principal.  5.13 (5.04)  5.40 (5.30)  Number of Years Teaching Prior to Principal  The number of years a principal spent teaching in the classroom prior to becoming a principal.  11.45 (11.49)  6.36 (6.37)  Control Variables  Northeast  School is located in the northeast.  0.16 (0.16)  0.37 (0.37)  Midwest  School is located in the midwest.  0.26 (0.26)  0.44 (0.44)  West  School is located in the west.  0.25 (0.25)  0.43 (0.43)  South  School is located in the south (Reference category).  0.34 (0.33)  0.47 (0.47)  Urban  School is located in an urban setting.  0.20 (0.20)  0.40 (0.40)  Suburban  School is located in a suburban setting.  0.31 (0.31)  0.46 (0.46)  Rural  School is located in a rural setting (Reference category).  0.49 (0.50)  0.50 (0.50)  Grades Served  Grades served in the school. Elementary schools are coded as 1. Combined and secondary schools are coded as 0.  0.55 (0.55)  0.50 (0.50)  Enrollment  The number of students enrolled in the school.  633.67 (642.86)  497.18 (504.32)  Percent Free Lunch  This variable is the percentage of students who are eligible to receive free lunch in the school (range: 0-100).  36.07 (37.16)  26.10 (26.41)  Percent White  Percent of White students enrolled in the school. (Reference category).  73.17 (72.47)  29.99 (29.99)  Percent Black  Percent of Black students enrolled in the school.  12.05 (11.91)  22.07 (21.73)  Percent Hispanic  Percent of Hispanic students enrolled in the school.  8.35 (9.05)  17.46 (18.05)  Percent Asian  Percent of Asian students enrolled in the school.  2.78 (2.79)  9.30 (8.93)  Percent American Indian  Percent of American Indian students enrolled in the school.  3.65 (3.52)  13.54 (13.23)  Teachers in a Tested Subject  The percent of teachers who responded within the school, and teach in a tested subject.  48.54 (49.04)  33.33 (33.17)  Duplicate School  Some schools apper more than once across all waves of the SASS data. I control for schools that appear more than once (coded as 1).  0.49 (0.49)  0.50 (0.50)  N = 34,940 (N=40,360)  Note: Data is from the National Center for Education Statistics (NCES) Schools and Staffing Survey (SASS). The sample uses all traditional schools in the 1987, 1990, 1993, 1999, 2003, 2007, and 2011 waves of SASS data that have valid information on my independent, dependent, and control variables. I exclude private, charter, alternative, and non-traditional schools. Analytic Strategy This analysis proceeds in two major steps. First, using all seven waves of data from 1987 to 2011, table 3 examines the general trends in micro-coupling over time. In model 1, I use a continuous measure of time to examine the overall effect of time on micro-level coupling. In model 2, I break the survey years into separate dummy variables in order to evaluate how each time period affects micro-level coupling. Second, using the pooled waves from 1987 to 2007 and clustered robust standard errors to account for the nesting of schools within states (ICC = 0.10), I use OLS regression to analyze how the tiered levels of the educational system contribute to micro-level coupling. I utilize a stepwise approach entering each of the tiered levels of interest—federal eras and state laws/accountability structures, local characteristics, and principal attributes—in a sequential fashion. Table 3. OLS Regression of Micro-Coupling on Federal Policy Eras   Model 1  Model 2  Socio-political factors      Survey waves (1987–2011)  0.007**    Year 1990    −0.050***  Year 1993    −0.065***  Year 1999    0.041**  Year 2003    −0.076***  Year 2007    −0.032*  Year 2011    0.049***  (Reference is 1987 (ANaR))      Controls      Location      Northeast  −0.099*  −0.104*  Midwest  −0.172***  −0.175***  West  −0.116**  −0.116**  (reference is South)      Setting      Urban  0.175***  0.179***  Suburban  0.124***  0.129***  (reference is Rural)      School demographics      Free lunch percentage  0.000  0.000  Black percentage  0.003***  0.003***  Hispanic percentage  0.002***  0.002***  Asian percentage  −0.002*  −0.002*  American Indian percentage  0.001  0.001  (reference is White percentage)      Enrollment  0.000***  0.000***  Elementary school  0.208***  0.215***  Percent of teachers in tested subject  0.002***  0.002***  Duplicate school  −0.019**  −0.011  (reference is Non-repeated school)      Constant  0.579***  0.631***  N = 40,360      R-squared  0.215  0.223    Model 1  Model 2  Socio-political factors      Survey waves (1987–2011)  0.007**    Year 1990    −0.050***  Year 1993    −0.065***  Year 1999    0.041**  Year 2003    −0.076***  Year 2007    −0.032*  Year 2011    0.049***  (Reference is 1987 (ANaR))      Controls      Location      Northeast  −0.099*  −0.104*  Midwest  −0.172***  −0.175***  West  −0.116**  −0.116**  (reference is South)      Setting      Urban  0.175***  0.179***  Suburban  0.124***  0.129***  (reference is Rural)      School demographics      Free lunch percentage  0.000  0.000  Black percentage  0.003***  0.003***  Hispanic percentage  0.002***  0.002***  Asian percentage  −0.002*  −0.002*  American Indian percentage  0.001  0.001  (reference is White percentage)      Enrollment  0.000***  0.000***  Elementary school  0.208***  0.215***  Percent of teachers in tested subject  0.002***  0.002***  Duplicate school  −0.019**  −0.011  (reference is Non-repeated school)      Constant  0.579***  0.631***  N = 40,360      R-squared  0.215  0.223  Note: Schools are clustered in states. Data are from the National Center for Education Statistics (NCES) Schools and Staffing Survey (SASS) and, following NCES convention, I have rounded sample size numbers to the nearest 10 in order to protect the identities of respondents. Given the large number of cases, models were examined using a random 10 percent subsample, and the results are substantively similar. The sample uses all traditional schools in the 1987, 1990, 1993, 1999, 2003, 2007, and 2011 waves of SASS data that have valid information on my independent, dependent, and control variables. I exclude private, charter, alternative, and non-traditional schools. *p < 0.05 **p < 0.01 ***p < 0.001 Due to data limitations, I am unable to present table 4 using data from the 2011 wave of data. In an  appendix, I include the 2011 survey year in models that largely replicate those presented in table 4. The meso-level coupling variable and the interaction term for gender and meso-level coupling are omitted from these models. Table 4. OLS Regression of Micro-Coupling on Federal, State, Local, and Principal Characteristics   Model 1  Model 2  Model 3  Model 4  Model 5  Socio-political factors            Year 1990  −0.052***  −0.050***  −0.054***  −0.048***  −0.054***  Year 1993  −0.075***  −0.070***  −0.077***  −0.058***  −0.077***  Year 1999 (IASA)  0.038*  0.042*  0.031  0.040*  0.031  Year 2003 (NCLB)  −0.076***  −0.077***  −0.092***  −0.118***  −0.092***  Year 2007  −0.026  −0.027  −0.042*  −0.073***  −0.042*  Reference is 1987 (ANaR)            State characteristics            Index of high-stakes testing  0.013  0.013  0.012  0.013  0.012  Republican state (reference is Democrat)  −0.022  −0.021  −0.020  −0.021  −0.020  Charter law  −0.005  −0.006  −0.007  −0.007  −0.007  Local characteristics            Tight meso-coupling (local govt to school relationship)    0.033***  0.031***  0.031***  0.018*  (Reference is 0–3)            Bonus    −0.012  −0.012  −0.005  −0.012  Principal’s characteristics            Female principal (reference is Male)      0.060***  0.055**  0.052***  Non-White principal (reference is White)      −0.003  −0.001  −0.004  Highest degree—Doctorate      0.003  0.002  0.003  (reference is All other degrees)            Principal tenure (in years)      −0.002***  −0.002***  −0.002***  Teaching experience (in years) prior to principal      0.000  0.000  0.000  Interaction terms            Female principal x Year 1990        −0.029    Female principal x Year 1993        −0.079***    Female principal x Year 1999        −0.027    Female principal x Year 2003        0.074**    Female principal x Year 2007        0.074**    Female principal x Tightest meso-coupling          0.036*  Controls            Location            Northeast  −0.099*  −0.099*  −0.098*  −0.098*  −0.098*  Midwest  −0.155**  −0.155**  −0.152**  −0.150**  −0.153**  West  −0.094  −0.093  −0.097  −0.095  −0.097  (reference is South)            Setting            Urban  0.184***  0.182***  0.175***  0.176***  0.175***  Suburban  0.128***  0.127***  0.122***  0.123***  0.122***  (reference is Rural)            School demographics            Free lunch percentage  −0.000  −0.000  −0.000  −0.000  −0.000  Black percentage  0.003***  0.003***  0.003***  0.003***  0.003***  Hispanic percentage  0.001***  0.001***  0.001**  0.001**  0.001**  Asian percentage  −0.002**  −0.002**  −0.002**  −0.002**  −0.002**  American Indian percentage  0.001  0.001  0.001  0.001  0.001  (reference is White percentage)            Enrollment  0.000***  0.000***  0.000***  0.000***  0.000***  Elementary school (reference is Secondary/Combined schools)  0.211***  0.211***  0.199***  0.198***  0.199***  Percent of teachers in tested subject  0.002***  0.002***  0.002***  0.002***  0.002***  Duplicate school  −0.002  −0.007  −0.005  0.000  −0.005  (reference is Non-repeated school)            Constant  0.628***  0.624***  0.636***  0.630***  0.638***  N = 34,940            R-squared  0.221  0.221  0.225  0.228  0.226    Model 1  Model 2  Model 3  Model 4  Model 5  Socio-political factors            Year 1990  −0.052***  −0.050***  −0.054***  −0.048***  −0.054***  Year 1993  −0.075***  −0.070***  −0.077***  −0.058***  −0.077***  Year 1999 (IASA)  0.038*  0.042*  0.031  0.040*  0.031  Year 2003 (NCLB)  −0.076***  −0.077***  −0.092***  −0.118***  −0.092***  Year 2007  −0.026  −0.027  −0.042*  −0.073***  −0.042*  Reference is 1987 (ANaR)            State characteristics            Index of high-stakes testing  0.013  0.013  0.012  0.013  0.012  Republican state (reference is Democrat)  −0.022  −0.021  −0.020  −0.021  −0.020  Charter law  −0.005  −0.006  −0.007  −0.007  −0.007  Local characteristics            Tight meso-coupling (local govt to school relationship)    0.033***  0.031***  0.031***  0.018*  (Reference is 0–3)            Bonus    −0.012  −0.012  −0.005  −0.012  Principal’s characteristics            Female principal (reference is Male)      0.060***  0.055**  0.052***  Non-White principal (reference is White)      −0.003  −0.001  −0.004  Highest degree—Doctorate      0.003  0.002  0.003  (reference is All other degrees)            Principal tenure (in years)      −0.002***  −0.002***  −0.002***  Teaching experience (in years) prior to principal      0.000  0.000  0.000  Interaction terms            Female principal x Year 1990        −0.029    Female principal x Year 1993        −0.079***    Female principal x Year 1999        −0.027    Female principal x Year 2003        0.074**    Female principal x Year 2007        0.074**    Female principal x Tightest meso-coupling          0.036*  Controls            Location            Northeast  −0.099*  −0.099*  −0.098*  −0.098*  −0.098*  Midwest  −0.155**  −0.155**  −0.152**  −0.150**  −0.153**  West  −0.094  −0.093  −0.097  −0.095  −0.097  (reference is South)            Setting            Urban  0.184***  0.182***  0.175***  0.176***  0.175***  Suburban  0.128***  0.127***  0.122***  0.123***  0.122***  (reference is Rural)            School demographics            Free lunch percentage  −0.000  −0.000  −0.000  −0.000  −0.000  Black percentage  0.003***  0.003***  0.003***  0.003***  0.003***  Hispanic percentage  0.001***  0.001***  0.001**  0.001**  0.001**  Asian percentage  −0.002**  −0.002**  −0.002**  −0.002**  −0.002**  American Indian percentage  0.001  0.001  0.001  0.001  0.001  (reference is White percentage)            Enrollment  0.000***  0.000***  0.000***  0.000***  0.000***  Elementary school (reference is Secondary/Combined schools)  0.211***  0.211***  0.199***  0.198***  0.199***  Percent of teachers in tested subject  0.002***  0.002***  0.002***  0.002***  0.002***  Duplicate school  −0.002  −0.007  −0.005  0.000  −0.005  (reference is Non-repeated school)            Constant  0.628***  0.624***  0.636***  0.630***  0.638***  N = 34,940            R-squared  0.221  0.221  0.225  0.228  0.226  Note: Schools are clustered in states. Data are from the National Center for Education Statistics (NCES) Schools and Staffing Survey (SASS) and, following NCES convention, I have rounded sample size numbers to the nearest 10 in order to protect the identities of respondents. Given the large number of cases, models were examined using a random 10 percent subsample, and the results are substantively similar. The sample uses all traditional schools in the 1987, 1990, 1993, 1999, 2003, and 2007 waves of SASS data that have valid information on my independent, dependent, and control variables. I exclude private, charter, alternative, and non-traditional schools. *p < 0.05 **p < 0.01 ***p < 0.001 Results The analyses presented in table 3 test the direct associations between federal policy era and micro-level coupling. Model 1 indicates a significant and positive relationship between time and micro-level coupling, suggesting that schools have tightened up coupling throughout the standards-based reform era, generally supporting previous findings (Hallett 2010). In model 2 of table 3, I include each survey wave as a dummy variable in order to explore the nuance of how the standards-based reform era affected micro-level coupling. This model shows that schools have not simply tightened their coupling over time; instead, the relationship is non-linear. I find a significant positive association between the 1999 and 2011 waves and micro-level coupling. Relative to 1987, the 1999 and 2011 waves are the only two eras with a positive association to micro-level coupling, suggesting periods of recoupling. The NCLB era (2003), relative to 1987, has a significant negative association with micro-coupling, and that association persists for the 2007 era, indicating a period of decoupling. Models 1 and 2 in table 3 demonstrate an interesting pattern across US public schools, and speak to the general trends in coupling. Current developments in educational research suggest a recoupling movement in public schools, and while model 1 depicts a small but steady increase in coupling across public schools, model 2 tells a more nuanced story about the relative effects of policy eras and illustrates a non-linear trend with schools experiencing decoupling, recoupling, decoupling, and finally recoupling again. Relevant federal reports and policies (i.e., A Nation at Risk, IASA, and NCLB) are noted in figure 1 to provide reference points. The trends in coupling help place the recent qualitative studies of recoupling and increased tighter coupling into context. Wald tests for model 2 in table 3 reveal that most contrasts between successive years are significantly different from one another (with 1990 to 1993 as exceptions). Therefore, the observed recoupling from 2003 to 2007 is a statistically significant change. The recoupling that is observed in 2007 does suggest that NCLB may have had longer-term effects on the organization of public schools. The recoupling trend continues from 2007 through 2011, and this is well past the implementation of NCLB. Figure 1. View largeDownload slide Graph of micro-coupling over time Note: Estimates based on model 2 in table 3. Figure 1. View largeDownload slide Graph of micro-coupling over time Note: Estimates based on model 2 in table 3. In table 4, I include the remaining tiers of the public education system. Model 1 of table 4 introduces state-level characteristics. None of the state characteristics are significantly associated with micro-level coupling. Recall that these school-level data are clustered by state and the statistical significance of the ICC denotes the importance of state-level characteristics. While state-level features modeled in table 4 are non-significant, 10 percent of the variance between schools is due to state-level differences not captured here. Model 1 of  table A1 (see the appendix) replicates this analysis and includes the 2011 wave of data. Model 2 steps in the meso-level structures, which include local-to-school coupling relationships and the presence of a performance-based merit pay or bonus system. As expected, tight meso-level coupling is positively associated with tighter micro-level coupling. Model 2 suggests that when meso-level coupling is tightest (i.e., in the fourth quartile), there is an increase of 0.033 on micro-level coupling. The micro-level coupling index ranges from 0 to 4, with a mean of 0.90, making 0.033 (approximately 7 percent of a standard deviation) a moderate increase in micro-level coupling. Most federal policy eras are significant in model 2 (2007 is the exception), suggesting that both federal and local elements shape school environments. Model 2 of  table A1 reproduces these models using the 2011 data (and excludes meso-level coupling). The results in  table A1 are substantively similar, and the survey year of 2011 is positively associated with tighter micro-level coupling. Model 3 brings together all levels of the public school system and explores the role of the principal. Schools with female principals are more tightly coupled at the micro-level. Although many other principal attributes are non-significant, the variable indicating principals’ tenure is negatively related to tight coupling. For every additional year a principal has served as a principal, there is a decrease of 0.002 in micro-coupling. Meso-level coupling remains positively significant across models. Most federal policy eras are also still significant in model 3 (except 1999); model 3 in  table A1 replicates these models using 2011 data (and excludes meso-level coupling) and shows a similar pattern, although the survey year of 2011 is not significant in this model. While model 3 conveys how all levels of socio-political factors directly affect schools, it does not adequately speak to the potential interaction between these levels of analysis. I ran two sets of interactions to explore how contexts matter for gendered effects. First, I ran an interaction between federal policy eras and female principals (model 4). Second, I ran an interaction between meso-level coupling and female principals (model 5). The interactions in model 4 of table 4 illustrate how federal policy eras vary by gendered leadership. Male and female principals are similar in many policy eras, but their management styles do vary in some contexts. Specifically, the interaction terms for the most recent waves (2003 and 2007) are significant and positive. Compared to their male counterparts, female principals in 2003 and 2007 are associated with tighter micro-level coupling. In model 4 of  table A1, I also model interactions that include the 2011 wave of data. The results are similar, but the interaction term for female principals in 2011 is non-significant. Figure 2 graphs the patterns of these interactions, and Wald tests reveal that most interaction terms are significantly different from one another (2003 to 2007 is the exception). I use a dotted line to graph the 2011 data because it is based on the analyses in model 4 of  table A1. Figure 2. View largeDownload slide Graph of interaction between principal’s gender and federal policy eras Note: The solid line from 1987 to 2007 uses data from model 4 in table 4. The dotted line from 2007 to 2011 uses data from model 4 in table A1 (see appendix). Figure 2. View largeDownload slide Graph of interaction between principal’s gender and federal policy eras Note: The solid line from 1987 to 2007 uses data from model 4 in table 4. The dotted line from 2007 to 2011 uses data from model 4 in table A1 (see appendix). Model 5 includes the interaction for meso-level coupling and gender, and the significant coefficient for the interaction means the association between meso-level coupling and micro-level coupling does vary by principals’ gender. Tight meso-level coupling is associated with a 0.054 point increase in micro-level coupling when the principal is female (bmeso + bmesoxfemale = 0.018 + 0.036 = 0.054), and only a 0.018 point increase in micro-level coupling when the principal is male. Figure 3 shows that compared to male principals, female principals are associated with tighter micro-level coupling directly, and female principals also strengthen the relationship between local-level coupling and micro-level coupling. A slope test revealed that the relationship between meso-level coupling and micro-level coupling is statistically significant for both male and female principals, but significantly stronger for female principals. I explored interactions between female principals and other school factors (i.e., race, urbanicity), but these were not significant and are thus not included here. Figure 3. View largeDownload slide Graph of interaction between principal’s gender and meso-level coupling Figure 3. View largeDownload slide Graph of interaction between principal’s gender and meso-level coupling The analyses reveal fascinating findings pertaining to school demographic characteristics. First, the general socioeconomic status, measured through free lunch eligibility, is not a significant predictor of tightly coupled schools, but the racial composition of the school is significantly associated with tighter coupling. As the percentage of Black and Hispanic students increases within the school, schools experience tighter coupling. Second, elementary schools (compared to secondary or combined schools) are positively associated with tighter coupling. Third, school setting is associated with coupling, and urban schools have tighter coupling compared to rural and suburban schools. Fourth, as the percentage of responding teachers who teach in tested subjects increases, we see tighter micro-level coupling. Finally, schools that appear multiple times in my data are not significantly affecting the results presented here. When I include the 2011 wave of data ( table A1 in the appendix), the findings are reproduced. These results tell a thought-provoking story about how multiple levels of the education system intersect to influence school coupling. Discussion and Conclusion These results challenge some of our taken-for-granted assumptions about school coupling and highlight the necessity of integrating literatures to provide a more complete picture. Empirically, many studies of coupling focus on the outcomes of tight coupling in the context of shifting policy environments. This study uses those analyses to provide a framework and a backdrop for evaluating the impetus for tighter coupling on a national level. I measure how different levels of the public education system—including gendered leadership structures within schools—contribute independently and interactively to tighter coupling within public schools. Tighter coupling is not merely a result of one hierarchical level exerting an overwhelming force on the interior structure of schools; rather, multiple levels must be considered simultaneously. These results suggest that a number of federal-level policy eras are more relevant than others. State-level factors had no significant relationship with school-level organization, despite the overwhelming credit often bestowed upon states and “states’ rights” in education policy. Local government coupling relationships (i.e., meso-level coupling) influence tighter coupling at the micro-level; stronger ties at the local level encourage tighter coupling within the school. Many principal characteristics are not significantly related to tightly coupled structures within schools. Principals’ tenure and gender both directly predict coupling. In fact, schools with female principals have tighter coupling than those with male principals. Finally, important intersections between context and gendered leadership help detail the comprehensive story of school coupling. This research uniquely addressed two competing ideas from organizational research. First, coupling scholars have found that creating or increasing accountability standards has engendered tighter coupling within schools. Second, neo-institutionalism posits that chaos and disorder will stimulate an environment of loose coupling within an organization. The standards-based reform era, specifically for the policy era of NCLB, provides an interesting situation where increased accountability and chaos occurred in the same time period. Federal policies structured around performance accountability gradually increased demands on public schools, culminating in the NCLB era. Unfortunately, NCLB is considered poorly executed through direction, funding, and mandates (Mathis 2003; Orlich 2004; Weeden 2005). Scholars criticize NCLB as chaotic and confusing for schools, principals, and teachers (Cochran-Smith and Lytle 2006; Darling-Hammond 2007a, 2007b; Le Floch, Taylor, and Thomsen 2006; Valli and Buese 2007). Organizational scholars discuss the implications of chaotic environments and predict that organizational actors will actively loosen structures of coupling when faced with chaotic institutional environments (Weick 1976). The presence of a federal policy that simultaneously produces increased accountability standards and chaos for schools creates a tension for this research and organizational research in general. My findings suggest that the transition into NCLB is negatively associated with tight coupling, and the confusion surrounding the federal policy was a stronger force than the pressure of accountability standards. This result is consequential, but it is possible that it is distinct to public schools. Public schools could be an exceptional organizational form, and not all organizations may respond similarly when faced with competing pressures of accountability and chaos. The general trend in micro-level coupling is non-linear, and while organizational scholars know organizations can change (Aldrich and Reuf 2006), it is unclear how often an internal organizational structure shifts. With regard to my findings on US public schools, coupling within schools has changed over time, but it has not simply steadily increased (or decreased) over time. The picture presented in this paper suggests movement, but it is also possible that change is temporary or never strays too far from an average. How changes in micro-level coupling affect daily life in schools is difficult to discern from these analyses, but are well supplemented by the qualitative research conducted during these policy eras. Considering this research in light of those studies helps show patterns in coupling, while simultaneously allowing us to reflect on the processes behind the changes in coupling. My analyses do not show large effects in any one time period. Indeed, all of the coefficients are smaller than 0.12, although across multiple time periods (e.g., from 1993 to 1999) the change in micro-level coupling is greater than 0.10. With a standard deviation of 0.47 for micro-coupling, a change slightly greater than 0.10 indicates just under one-quarter of a standard deviation change from 1993 to 1999. The exact meaning behind one-fifth to one-quarter of a standard deviation is not evidenced in the models presented here, but several studies were conducted during the highest peak illustrated in figure 1 (see, for instance, Coburn 2004; Diamond 2007; Hallett 2010), and this research helps exhibit what tighter coupling looks like in teachers’ day-to-day lives. These studies help put my nationally representative findings into context, and imply that even what appears to be a small to moderate shift from looser to tighter coupling within schools does effect change within their walls. If policymakers hope principals and teachers are forming tight linkages within schools, then the non-linear relationship between federal policy eras and micro-coupling presents a unique dilemma for shaping future federal-level policies. There was a significant loosening in coupling from 1999 to 2003, which may surprise educational policymakers. But the latter part of the NCLB era does depict an increase in micro-level coupling (from 2003), and it is possible that schools rebounded from the initial shock of NCLB mandates. Another reauthorization of ESEA means we are moving into a new policy era under the Every Student Succeeds Act (ESSA), and public schools may undergo another shift in coupling. Unlike IASA and NCLB, ESSA sends power back to the states to develop their own standards of accountability. Consequently, federal policy may become less important while school demographics and gendered leadership take center stage in molding schools’ micro-level coupling. Regardless, paying close attention to the trends of micro-coupling across public schools will be important for those interested in the organizational structure of schools. As principals gain more experience, schools become more loosely coupled. This finding implies that experience leads to looser coupling in schools. It is possible that more experienced principals recognize their teachers’ desire for loose coupling, and allow them more freedom. On average, female principals spend more time in the classroom prior to becoming principals, but female principals promote tighter couplings. Gender also intersects with federal policy eras in compelling ways. In recent waves, under NCLB, female principals were more likely than their male colleagues to manage tightly coupled schools. Furthermore, when female principals administrate in schools with a strong local government influence, the relationship to tight coupling is strengthened. These intersections emphasize how different levels of the school system work in concert with one another. The relationship between female principals and tighter coupling could signify that females are seeking legitimacy in their authority by enacting policies within the walls of their schools, especially if female principals are also seeking legitimacy from district officials when the district is tightly coupled to the school (meso-level coupling). Indeed, female principals have traditionally managed their schools using more participative strategies (Shakeshaft 1987). Tight meso-level coupling could prompt the district to hire a female principal who will actively impose policies within the school, although qualitative studies highlight how tighter couplings may undermine the legitimacy of the principal (with teachers) (Hallett 2007b, 2010). These are two possibilities that should be explored in future research. My analyses suggest a persistent role of race, but not class, in school coupling. The overarching goals of federal policies underscore the need to close the achievement gap for poor students and racial minorities, which could help explain the racial differences found here, but perhaps generate more questions regarding schools that serve poor students. These findings warrant more exploration, and future research should consider the processes under which principals and teachers shift their organizational structure as racial and socio-economic demographics shift. Longitudinal data with the same schools over time would be especially useful for scrutinizing this relationship. Perhaps unsurprisingly, urban schools have the tightest school-level coupling. Urban schools are often targeted in federal- and state-level policies. Recent qualitative studies on tighter couplings were conducted in urban schools and/or schools with higher proportions of minority students, and the findings presented here suggest that demographic characteristics are related to school-level coupling on a national level. These results are intriguing and merit further investigation, especially if federal- and/or local-level policies disproportionately affect some schools more than others. Finally, this study considers very specific dimensions of principal-teacher relationships, and coupling may tighten on some dimensions while loosening on others. Future research should consider other aspects of the principal-teacher relationship in analyzing how couplings exist and change within schools. While this work focuses on how couplings develop within schools, processes of coupling are embedded in other types of organizations, and this research expands our theoretical understanding of coupling. The public education system provides one type of organization to empirically test theories of coupling and neo-institutionalism, but schools are not the only organizations that experience these organizational processes. Similarly, the role of gender in positions of leadership is a critical component for constructing couplings. Understanding broader theoretical issues through the public education system will help scholars advance perspectives of coupling. Notes 1 Other characteristics of principals matter for management style—race could be an important factor, but teaching is an interesting organizational setting when it comes to gender and management. The overwhelming white demographic of teachers (86 percent) and principals (87 percent) makes understanding racial differences difficult in quantitative analysis. 2 The SASS data is a restricted-use dataset; this number is rounded in order to protect confidentiality. The number 34,940 is not the precise number of observations in this analysis, but it is the number reported in tables. 3 This indicates the percentage of teachers in these data, not the total percent of teachers in the school. Appendix Table A1. OLS Regression of Micro Coupling on Federal, State, Local, and Principal Characteristics   Model 1  Model 2  Model 3  Model 4  Socio-political factors          Year 1990  −0.052***  −0.052***  −0.057***  −0.049***  Year 1993  −0.075***  −0.074***  −0.082***  −0.061***  Year 1999  0.035  0.036  0.024  0.035  Year 2003  −0.076**  −0.075**  −0.091***  −0.117***  Year 2007  −0.031  −0.030  −0.046*  −0.075***  Year 2011  0.043*  0.043*  0.027  0.017  (Reference is 1987 (ANaR))          State characteristics          Index of high-stakes testing  0.011  0.011  0.011  0.011  Republican state (reference is Democrat)  −0.022  −0.022  −0.021  −0.021  Charter law  −0.002  −0.002  −0.003  −0.004  Local characteristics          Bonus    −0.011  −0.011  −0.010  Principal’s characteristics          Female principal (reference is Male)      0.061***  0.053**  Non-White principal (reference is White)      0.000  0.002  Highest degree—Doctorate      0.002  0.002  (reference is All other degrees)          Principal tenure (in years)      −0.003***  −0.003***  Teaching experience (in years) prior to principal      0.000  0.000  Interaction terms          Female principal x Year 1990        −0.031  Female principal x Year 1993        −0.078***  Female principal x Year 1999        −0.028  Female principal x Year 2003        0.073**  Female principal x Year 2007        0.073**  Female principal x Year 2011        0.029  Controls          Location          Northeast  −0.106*  −0.107*  −0.105*  −0.105*  Midwest  −0.166***  −0.166***  −0.163***  −0.162***  West  −0.105*  −0.106*  −0.109*  −0.107*  (reference is South)          Setting          Urban  0.180***  0.180***  0.173***  0.174***  Suburban  0.125***  0.125***  0.120***  0.121***  (reference is Rural)          School demographics          Free lunch percentage  0.000  0.000  0.000  0.000  Black percentage  0.003***  0.003***  0.003***  0.003***  Hispanic percentage  0.002***  0.002***  0.001***  0.001***  Asian percentage  −0.002**  −0.002**  −0.002**  −0.002**  American Indian percentage  0.001  0.001  0.001  0.001  (reference is White percentage)          Enrollment  0.000***  0.000***  0.000***  0.000***  Elementary school (reference is Secondary/Combined schools)  0.214***  0.214***  0.203***  0.202***  Percent of teachers in tested subject  0.002***  0.002***  0.002***  0.002***  Duplicate school  −0.007  −0.007  −0.005  −0.004  (Reference is Non-repeated school)          Constant  0.623***  0.623***  0.637***  0.635***  N = 40,360          R-squared  0.224  0.224  0.228  0.231    Model 1  Model 2  Model 3  Model 4  Socio-political factors          Year 1990  −0.052***  −0.052***  −0.057***  −0.049***  Year 1993  −0.075***  −0.074***  −0.082***  −0.061***  Year 1999  0.035  0.036  0.024  0.035  Year 2003  −0.076**  −0.075**  −0.091***  −0.117***  Year 2007  −0.031  −0.030  −0.046*  −0.075***  Year 2011  0.043*  0.043*  0.027  0.017  (Reference is 1987 (ANaR))          State characteristics          Index of high-stakes testing  0.011  0.011  0.011  0.011  Republican state (reference is Democrat)  −0.022  −0.022  −0.021  −0.021  Charter law  −0.002  −0.002  −0.003  −0.004  Local characteristics          Bonus    −0.011  −0.011  −0.010  Principal’s characteristics          Female principal (reference is Male)      0.061***  0.053**  Non-White principal (reference is White)      0.000  0.002  Highest degree—Doctorate      0.002  0.002  (reference is All other degrees)          Principal tenure (in years)      −0.003***  −0.003***  Teaching experience (in years) prior to principal      0.000  0.000  Interaction terms          Female principal x Year 1990        −0.031  Female principal x Year 1993        −0.078***  Female principal x Year 1999        −0.028  Female principal x Year 2003        0.073**  Female principal x Year 2007        0.073**  Female principal x Year 2011        0.029  Controls          Location          Northeast  −0.106*  −0.107*  −0.105*  −0.105*  Midwest  −0.166***  −0.166***  −0.163***  −0.162***  West  −0.105*  −0.106*  −0.109*  −0.107*  (reference is South)          Setting          Urban  0.180***  0.180***  0.173***  0.174***  Suburban  0.125***  0.125***  0.120***  0.121***  (reference is Rural)          School demographics          Free lunch percentage  0.000  0.000  0.000  0.000  Black percentage  0.003***  0.003***  0.003***  0.003***  Hispanic percentage  0.002***  0.002***  0.001***  0.001***  Asian percentage  −0.002**  −0.002**  −0.002**  −0.002**  American Indian percentage  0.001  0.001  0.001  0.001  (reference is White percentage)          Enrollment  0.000***  0.000***  0.000***  0.000***  Elementary school (reference is Secondary/Combined schools)  0.214***  0.214***  0.203***  0.202***  Percent of teachers in tested subject  0.002***  0.002***  0.002***  0.002***  Duplicate school  −0.007  −0.007  −0.005  −0.004  (Reference is Non-repeated school)          Constant  0.623***  0.623***  0.637***  0.635***  N = 40,360          R-squared  0.224  0.224  0.228  0.231  Note: Schools are clustered in states. Data are from the National Center for Education Statistics (NCES) Schools and Staffing Survey (SASS) and, following NCES convention, I have rounded sample size numbers to the nearest 10 in order to protect the identities of respondents. Given the large number of cases, models were examined using a random 10 percent subsample, and the results are substantively similar. The sample uses all traditional schools in the 1987, 1990, 1993, 1999, 2003, 2007, and 2011 waves of SASS data that have valid information on my independent, dependent, and control variables. I exclude private, charter, alternative, and non-traditional schools. *p < 0.05 **p < 0.01 ***p < 0.001 About the Author Maria Paino is an Assistant Professor in the Department of Sociology, Anthropology, Social Work, and Criminal Justice at Oakland University. Her research focuses on inequalities and organizational processes particularly in educational and clinical contexts. References Acker, Joan. 1990. “ Hierarchies, Jobs, Bodies: A Theory of Gendered Organizations.” Gender & Society  4( 2): 139– 58. Google Scholar CrossRef Search ADS   Aldrich, Howard E., and Martin Reuf. 2006. Organizations Evolving . London: Sage. Bidwell, Charles E. 2001. “ Analyzing Schools as Organizations: Long Term Permanence and Short Term Change.” Sociology of Education  ( Extra Issue): 100– 114. Bossert, Steven T., David C. Dwyer, Brian Rowan, and Ginny V. Lee. 1982. “ The Instructional Management Role of the Principal.” Educational Administration Quarterly  18( 3): 34– 64. Google Scholar CrossRef Search ADS   Carnoy, Martin, and Susanna Loeb. 2002. “ Does External Accountability Affect Student Outcomes? A Cross-State Analysis.” Educational Evaluation and Policy Analysis  24( 4): 305– 31. Google Scholar CrossRef Search ADS   Charters, W. W. Jr., and T. D. Jovick. 1981. “The Gender of Principals and Principal/Teacher Relations in Elementary Schools.” In Educational Policy and Management: Sex Differentials , edited by P. A. Schmuck, W. W. Charters Jr., and R. O. Carlson, 307– 331. New York: Academic Press. Coburn, Cynthia E. 2004. “Beyond Decoupling: Rethinking the Relationship between the Institutional Environment and the Classroom.” Sociology of Education  77( 3): 211– 44. Google Scholar CrossRef Search ADS   Cochran-Smith, Marilyn, and Susan L. Lytle. 2006. “ Troubling Images of Teaching in No Child Left Behind.” Harvard Educational Review  76( 4): 668– 97. Google Scholar CrossRef Search ADS   Cohen, Michael D., and James G. March. 1974. Leadership and Ambiguity . New York: McGraw-Hill. Darling-Hammond, Linda. 2007a. “ Race, Inequality and Educational Accountability: The Irony of ‘No Child Left Behind.’” Race Ethnicity and Education  10( 3): 245– 60. Google Scholar CrossRef Search ADS   Darling-Hammond, Linda. 2007b. “Evaluating ‘No Child Left Behind.’” The Nation 21(May):1–7. Davies, Scott, Linda Quirke, and Janice Aurini. 2006. “The New Institutionalism Goes to the Market: The Challenge of Rapid Growth in Private K-12 Education.” In The New Institutionalism in Education , edited by H. D. Meyer and B. Rowan, 103– 22. Albany, NY: SUNY Press. Desimone, Laura. 2013. “ Teacher and Administrator Responses to Standards-Based Reform.” Teachers College Record  115( 8): 1– 53. Diamond, John B. 2007. “ Where the Rubber Meets the Road: Rethinking the Connection between High-Stakes Testing Policy and Classroom Instruction.” Sociology of Education  80( 4): 285– 313. Google Scholar CrossRef Search ADS   Diamond, John B. 2012. “ Accountability Policy, School Organization, and Classroom Practice: Partial Recoupling and Educational Opportunity.” Education and Urban Society  44( 2): 151– 82. Google Scholar CrossRef Search ADS   Eagly, Alice H., and Blair T. Johnson. 1990. “ Gender and Leadership Style: A Meta-Analysis.” Psychological Bulletin  108: 233– 56. Google Scholar CrossRef Search ADS   Eagly, Alice H., and Steven J. Karau. 2002. “ Role Congruity Theory of Prejudice Towards Female Leaders.” Psychological Review  109( 3): 573– 98. Google Scholar CrossRef Search ADS PubMed  Elmore, Richard F., Charles Abelman, and Susan H. Furman. 1996. “The New Accountability in State Education Reform: From Process to Performance.” In Holding Schools Accountable , edited by H. F. Ladd, 65– 98. Washington, DC: Brookings Institution. Fauth, Gloria C. 1984. “ Women in Educational Administration: A Research Profile.” Educational Forum  49( 1): 65– 79. Google Scholar CrossRef Search ADS   Firestone, W. A. 1985. “The Study of Loose Coupling: Problems, Progress, and Prospects.” In Research in Sociology of Education and Socialization , vol. 5, edited by A. Kerckhoff, 3– 30. Greenwich, CT: JAI Press. Gamoran, Adam, and Robert Dreeben. 1986. “ Coupling and Control in Educational Organizations.” Administrative Science Quarterly  31: 612– 32. Google Scholar CrossRef Search ADS   Grissom, Jason A., Demetra Kalogrides, and Susanna Loeb. 2013. “Strategic Staffing: Examining the Class Assignments of Teachers and Students in Tested and Untested Grades and Subjects.” Paper presented at the Association for Education Finance and Policy, March 14, New Orleans, LA. Hallett, Tim. 2007a. “ Between Deference and Distinction: Interaction Ritual through Symbolic Power in an Educational Institution.” Social Psychology Quarterly  70( 2): 148– 71. Google Scholar CrossRef Search ADS   Hallett, Tim. 2007b. “The Leadership Struggle: The Case of Costen Elementary School.” In Distributed Leadership in Practice , edited by J. Spillane and J. B. Diamond, 85– 105. New York: Teachers College Press. Hallett, Tim. 2010. “ The Myth Incarnate: Recoupling Processes, Turmoil, and Inhabited Institutions in an Urban Elementary School.” American Sociological Review  75( 1), 52– 74. Google Scholar CrossRef Search ADS   Hanson, Mark. 1977. “ Beyond the Bureaucratic Model: A Study of Power and Autonomy in Educational Decision-Making.” Interchange  7( 2): 27– 38. Google Scholar CrossRef Search ADS   Ingersoll, Richard M. 1996. “ Teachers’ Decision-Making Power and School Conflict.” Sociology of Education  69( 2): 159– 76. Google Scholar CrossRef Search ADS   Ingersoll, Richard M. 2003. Who Controls Teachers’ Work? Power and Accountability in American Schools . Cambridge, MA: Harvard University Press. Jennings, Jennifer L. 2010. “ School Choice or Schools’ Choice? Managing in an Era of Accountability.” Sociology of Education  83( 3): 227– 47. Google Scholar CrossRef Search ADS   Johnson, Shirley, Steve Busch, and John R. Slate. 2008. “Leadership Behaviors of School Administrators: Do Men and Women Differ?” International Journal of Educational Leadership Preparation 3(3): 1– 11. Le Floch, Kerstin Carlson, James E. Taylor, and Kerri Thomsen. 2006. “ Implications of Nclb Accountability for Comprehensive School Reform.” Journal of Education for Students Placed at Risk  11( 3– 4): 353– 66. Google Scholar CrossRef Search ADS   Lee, Valerie E., Robert F. Dedrick, and Julia B. Smith. 1991. “ The Effect of the Social Organization of Schools on Teachers’ Efficacy and Satisfaction.” Sociology of Education  64( 3): 190– 208. Google Scholar CrossRef Search ADS   Lee, Valerie E., Julia B. Smith, and Madalyn Cioci. 1993. “ Teachers and Principals: Gender-Related Perceptions of Leadership and Power in Secondary Schools.” Educational Evaluation and Policy Analysis  15( 2): 153– 80. Google Scholar CrossRef Search ADS   Lortie, Dan C. 2002. Schoolteacher: A Sociological Study . Chicago: University of Chicago Press. March, James G., and Johan P. Olsen. 1976. Ambiguity and Choice in Organizations . Bergen: Universitetsforlaget. Mathis, William J. 2003. “ No Child Left Behind: Costs and Benefits.” The Phi Delta Kappan  84( 9): 679– 86. Google Scholar CrossRef Search ADS   McDermott, Kathryn A. 2011. High Stakes Reform: The Politics of Educational Accountability . Washington DC: Georgetown University Press. Meyer, John W., and Brian Rowan. 1977. “ Institutionalized Organizations: Formal Structure as Myth and Ceremony.” American Journal of Sociology  83: 340– 63. Google Scholar CrossRef Search ADS   Meyer, John W., and Brian Rowan. 1978. “The Structure of Educational Organizations.” In Environments and Organizations , edited by M. W. Meyer, 78– 109. San Francisco: Jossey-Bass. National Center for Education Statistics. 1994. “Public and Private School Principals: Are There Too Few Women?” Washington, DC: US Department of Education. National Center for Education Statistics. 2013. “Characteristics of Public and Private Elementary and Secondary School Principals in the United States.” Washington, DC: US Department of Education. Office of the Federal Register. 2013. “U.S. Electoral College.” Washington, DC: US National Archives and Records Administration. http://www.archives.gov/federal-register/electoral-college/about.html. Orlich, Donald C. 2004. “ No Child Left Behind: An Illogical Accountability Model.” The Clearing House  78( 1): 6– 11. Google Scholar CrossRef Search ADS   Orton, J. Douglas, and Karl E. Weick. 1990. “ Loosely Coupled Systems: A Reconceptualization.” Academy of Management Review  15( 2): 203– 23. Pfeffer, Jeffrey. 1981. Power in Organizations . Boston: Pitman. Pierce, Jennifer. 1995. Gender Trials . Berkeley: University of California Press. Pitner, Nancy J. 1981. “Hormones and Harems: Are the Activities of Superintending Different for a Woman?” In Educational Policy and Management: Sex Differentials , edited by P. A. Schmuck, W. W. Charters Jr., and R. O. Carlson, 273– 295. New York: Academic Press. Price, Heather. 2012. “ Principal-Teacher Interactions: How Affective Relationships Shape Principal and Teacher Attitudes.” Educational Administration Quarterly  48( 1): 39– 85. Google Scholar CrossRef Search ADS   Renzulli, Linda A., and Vincent J. Roscigno. 2005. “ Charter School Policy, Implementation, and Diffusion across the United States.” Sociology of Education  78( 4): 344– 65. Google Scholar CrossRef Search ADS   Renzulli, Linda A., Heather Macpherson Parrott, and Irenee R. Beattie. 2011. “ Racial Mismatch and School Type: Teacher Satisfaction and Retention in Charter and Traditional Public Schools.” Sociology of Education  84( 1): 23– 48. Google Scholar CrossRef Search ADS   Rowan, Brian. 1981. “The Effects of Institutionalized Rules on Administrators.” In Organizational Behavior in Schools and School Districts , edited by S. B. Bacharach, 47– 75. New York: Praeger. Rowan, Brian. 1990. “ Commitment and Control: Alternative Strategies for the Organizational Design of Schools.” Review of Research in Education  16: 359– 89. Sauder, Michael, and Wendy Nelson Espeland. 2009. “ The Discipline of Rankings: Tight Coupling and Organizational Change.” American Sociological Review  74: 63– 82. Google Scholar CrossRef Search ADS   Seyfarth, John T., and William A. Bost. 1982. “ Prejudice in the Principal’s Office: Administrator Attitudes toward Older Teachers.” High School Journal  65( 5): 157– 62. Shakeshaft, Charol. 1987. Women in Educational Administration . Newbury Park, CA: Sage. Spillane, James P., and Patricia Burch. 2006. “The Institutional Environment and Instructional Practice: Changing Patterns of Guidance and Control in Public Education.” In The New Institutionalism in Education , edited by H.-D. Meyer and B. Rowan, 87– 102. Albany: State University of New York Press. Spillane, James P., John B. Diamond, Patricia Burch, Tim Hallett, Loyiso Jita, and Jennifer Zoltners. 2002. “ Managing in the Middle: School Leaders and the Enactment of Accountability Policy.” Educational Policy  16( 5): 731– 62. Google Scholar CrossRef Search ADS   Spillane, James P., Tim Hallett, and John B. Diamond. 2003. “ Forms of Capital and the Construction of Leadership: Instructional Leadership in Urban Elementary Schools.” Sociology of Education  76( 1): 1– 17. Google Scholar CrossRef Search ADS   Spillane, James P., Leigh Mesler Parise, and Jennifer Zoltners Sherer. 2011. “ Organizational Routines as Coupling Mechanisms: Policy, School Administration, and the Technical Core.” American Educational Research Journal  48( 3): 586– 619. Google Scholar CrossRef Search ADS   Tyler, Tom R. 2006. “ Psychological Perspectives on Legitimacy and Legitimation.” Annual Review of Psychology  57: 375– 400. Google Scholar CrossRef Search ADS PubMed  US Department of Education. 2013. “States Granted Waivers from No Child Left Behind Allowed to Reapply for Renewal for 2014 and 2015 School Years.” Washington, DC. Valli, Linda, and Daria Buese. 2007. “ The Changing Roles of Teachers in an Era of High-Stakes Accountability.” American Educational Research Journal  44( 3): 519– 58. Google Scholar CrossRef Search ADS   Weeden, L. Darnell. 2005. “Essay: Does the No Child Left Behind Law (NCLBA) Burden the States as an Unfunded Mandate under Federal Law [article]." Thurgood Marshall Law Review 31( 2): 239– 252. Weick, Karl. 1976. “ Educational Organizations as Loosely Coupled Systems.” Administrative Science Quarterly  21: 1– 19. Google Scholar CrossRef Search ADS   Weiss, Carol H., and Joseph Cambone. 1994. “ Principals, Shared Decision Making, and School Reform.” Educational Evaluation and Policy Analysis  16( 3): 287– 301. doi:10.2307/1164401. Google Scholar CrossRef Search ADS   Williams, Christine L. 1993. Doing “Women’s Work”: Men in Nontraditional Occupations . Newbury Park: Sage. Google Scholar CrossRef Search ADS   Wingfield, Adia Harvey. 2009. “ Racializing the Glass Escalator: Reconsidering Men’s Experiences with Women’s Work.” Gender and Society  23( 1): 5– 26. doi:10.2307/20676747. Google Scholar CrossRef Search ADS   Young, Viki M. 2006. “ Teachers’ Use of Data: Loose Coupling, Agenda Setting, and Team Norms.” American Journal of Education  112( August): 521– 48. Google Scholar CrossRef Search ADS   Zelditch, Morris Jr. 2001. “Theories of Legitimacy.” In The Psychology of Legitimacy: Emerging Perspectives on Ideology, Justice, and Intergroup Relations , edited by J. Jost and B. Major, 33– 53. Cambridge: Cambridge University Press. Author notes I would like to thank Linda Renzulli, Ashley Barr, Elizabeth DeBray, Jeremy Reynolds, Rebecca Boylan, and Dennis Condron for their comments on earlier drafts. Please feel free to contact me at paino@oakland.edu. © The Author 2017. Published by Oxford University Press on behalf of the University of North Carolina at Chapel Hill. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Social Forces Oxford University Press

From Policies to Principals: Tiered Influences on School-Level Coupling

Loading next page...
 
/lp/ou_press/from-policies-to-principals-tiered-influences-on-school-level-coupling-NlWSmgPK6l
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press on behalf of the University of North Carolina at Chapel Hill. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
0037-7732
eISSN
1534-7605
D.O.I.
10.1093/sf/sox075
Publisher site
See Article on Publisher Site

Abstract

Abstract For decades, educational scholars have claimed that public schools are loosely coupled organizations, but research does not fully address how schools create the coupling structure. Recent federal policies have centralized the public education system, but state laws, districts, and principals still play a prominent role in shaping the organization of schools. In this paper, I investigate how the multiple tiers of the public education system influence the coupling structures of schools. Utilizing all seven waves of the Schools and Staffing Survey (SASS), I rely on OLS regression to analyze how couplings emerge within schools. Findings suggest three major influences on coupling within schools. First, federal policies have had a non-linear effect on school-level coupling. Second, the relationship between the district and the school affects school-level coupling. Finally, principals play a key role in shaping the coupling within the school, and coupling is a gendered process. Sociologists have described schools as loosely coupled organizations for decades, meaning the subparts of schools are linked and responsive yet largely autonomous (Bidwell 2001; March and Olsen 1976; Orton and Weick 1990; Rowan 1990; Weick 1976). Simultaneously, schools are gendered organizations that largely employ white women in middle-class semi-professions (Acker 1990; Hallett 2007a; Pierce 1995; Williams 1993). Males traditionally assumed leadership roles in schools, but recently the proportion of female principals has risen. This demographic shift makes gendered differences in leadership among principals increasingly salient. To date, these separate threads of research on organizational coupling and gendered leadership in schools have not spoken to each other. As accountability policies and national angst about the quality of our schools becomes a public concern, the work of teachers and their managers (principals) collides with the way schools are organized. It is necessary to integrate these literatures to better understand when couplings change, and this requires an approach that looks at the multiple levels of the public school system. In recent decades, accountability in the US public education system has increased. State and local governments remain important forces for public schools, but large-scale federal changes may drive changes in coupling between principals and teachers. Standards-based reform increased accountability demands on public schools by creating a more centralized federal structure (McDermott 2011). Federal policies have the potential to shift the structure of public schools, but schools are managed by principals, and principals must enact the federal policies in order for the policies to provoke change within schools (Hallett 2007a; Jennings 2010; Spillane et al. 2002; Spillane, Parise, and Sherer 2011). Principals play a key role in negotiating the relationships between policies and teachers (Lee, Smith, and Cioci 1993; Price 2012). As policies enforce more accountability, testing, and oversight, principals can influence how these policies are adopted and navigated by teachers (Diamond 2007). Over this same time period, women started to break the glass ceiling and gain jobs as principals—from 21 percent in the mid-1980s (NCES 1994) to 51 percent in recent years (NCES 2013). The increase in female principals demands we pay attention to the literature on gendered leadership, and opens up new avenues of inquiry about the intersection between different levels of the school system that affect the inner workings of schools. Research on organizations and leadership tells us that men and women will lead differently (Bossert et al. 1982; Fauth 1984; Johnson, Busch, and Slate 2008; Lee, Smith, and Cioci 1993), and there are noteworthy differences between male and female principals’ leadership and management styles (Lee, Smith, and Cioci 1993; Price 2012). It is plausible that these female principals will enact coupling structures at the school level differently from their male colleagues. How do the multiple levels of the public school system affect school-level coupling? And in what context—federal, state, and/or local—does gender matter for coupling? Coupling is a central issue to the sociology of education, and I revisit foundational questions from a new conceptual and methodological approach. I combine two major literatures that address how schools are organized—coupling and gendered leadership. This is the first large-scale quantitative study to examine coupling from the federal level to the schoolhouse door. My analyses of nationally representative data provide a more nuanced picture of what coupling looks like in public schools across two and a half decades and two major policy eras. Coupling in Schools Coupling refers to how closely formal organizational structures (e.g., policy) are related to the real technical activities occurring within the organization (Weick 1976). Schools are loosely coupled when principals and teachers are tied to one another, but retain a relatively great degree of independence. For instance, when teachers have autonomy and control over their classrooms, and are not tightly connected to their principals, the schools are more loosely coupled. But if principals exert a great deal of influence over teachers’ classrooms, then the schools are more tightly coupled. Coupling is not a binary, however, where schools are either loosely coupled or tightly coupled. Rather, scholars have highlighted the dialectical nature of leadership practice within schools, which implies that coupling is not unidimensional. This understanding allows us to focus on how some dimensions are more tightly coupled, while others are more loosely coupled. In fact, work has suggested that coupling on some dimensions, such as instructional content (especially in math and language arts), is tighter than on other dimensions, such as pedagogical decisions (Diamond 2007). Weick (1976) labels schools as loosely coupled organizations and explains how loose coupling protects individuals and subunits of an organization during times of uncertainty. It may not pervade the entire organization or affect the organization at all times, however, and some areas of the organization may have tighter couplings in particular eras while other areas of the organization have looser couplings during other policy eras. Neo-institutionalists argue that schools attempt to maintain legitimacy by engaging in ceremonial displays of compliance, but do not alter the actual activities in classrooms (Meyer and Rowan 1977, 1978). These conceptualizations help organizational scholars make sense of the fact that schools’ structures remained largely unchanged despite large-scale policy changes. A review of the literature differentiates between eight types of loose coupling: “loose coupling (1) among individuals, (2) among subunits, (3) among organizations, (4) between hierarchical levels, (5) between organizations and environments, (6) among ideas, (7) between activities, and (8) between intentions and actions” (Orton and Weick 1990, 208, numbers added). Here, I focus on loose coupling in schools’ hierarchical levels (e.g., between the principal and teachers in schools) (Firestone 1985; Gamoran and Dreeben 1986). When educational scholars assume that schools are loosely coupled, they are often referring to the relationship between principals and teachers, but other aspects of a school may be tightly coupled even when principals and teachers are loosely coupled. Weick (1976) also points out that loose coupling in one arena of an organization suggests tight coupling elsewhere. Previous work on hierarchical levels does not fully capture all the levels of our public school system. I build upon this literature, focusing on hierarchical coupling at three distinct levels. First, I define micro-level coupling as the formal relationship between the principal and the teachers within the school. Second, I define meso-level coupling as the formal relationship between the school district and the school. Finally, I consider macro-level coupling as the relationship between accountability policies and the school. Macro structures include federal and state policies within this level of the tiered school system. My conceptualization of coupling allows me to consider the various levels of the school system and the roles they play in shaping micro-level coupling. By taking multiple levels into account, we begin to understand how some levels may tighten coupling, while other levels loosen coupling, even within the same time period. Since the 1970s, studies have suggested that schools are loosely coupled and have been throughout history (Bidwell 2001; Gamoran and Dreeben 1986; March and Olsen 1976; Orton and Weick 1990; Rowan 1990; Weick 1976). In general, scholars underscore the principal-to-teacher relationship (i.e., micro-coupling), and indicate that principals exclude themselves from many day-to-day teaching and learning activities. Teachers oversee the lion’s share of instructional decisions (1978). Teachers do not have much control over school-level decisions (Lee, Dedrick, and Smith 1991; Renzulli, Parrott, and Beattie 2011), but they do strongly desire and seek to preserve autonomy over classroom affairs (Hanson 1977; Lortie 2002). The emphasis on autonomy, control, and input over classroom behavior signifies the sizable importance of relationships between teachers and principals in the loose coupling discourse. Over the past twenty years, various studies have suggested a surprising trend from loose to tight coupling (recoupling) within schools. Specifically, teachers report a decline in autonomy over classroom and school-related decisions (Coburn 2004; Diamond 2007, 2012; Hallett 2010; Ingersoll 2003; Young 2006). For example, teachers report “turmoil” when their administration replaces a loosely coupled system with tight coupling in response to policy changes (Hallett 2010). Teachers favor loosely coupled school environments when faced with increased institutional pressures (Coburn 2004; Ingersoll 2003), but studies have not fully investigated how the combined role of macro- (federal and state) and meso-level (local) institutional structures are catalysts for tighter or looser coupling on a national scale. I contribute to the discussion of coupling in organizations by using a tiered approach to understand how schools are coupled over several policy eras. Empirical studies of coupling have focused on the consequences of loose or tight coupling and have explored their sources. Research underscores why loose coupling is ideal for organizations (Cohen and March 1974; March and Olsen 1976; Meyer and Rowan 1977; Rowan 1981; Sauder and Espeland 2009), why teachers prefer loose coupling (Ingersoll 2003; Lortie 2002), and how teachers attempt to regain a loosely coupled structure when faced with threats to loose coupling (Coburn 2004). The impetus behind organizational couplings and the general trends of couplings at a national level is necessary, and the more recent findings on tighter coupling and recoupling motivate this study of how various tiers of the school system affect coupling. In order to fully understand how various tiers of the public school system affect coupling within public schools, we need to attend to accountability structures and district-school relationships. As noted above, we also need to engage the literature on gendered leadership. I use multiple levels of the educational system—federal, state, local, and principal—to offer a comprehensive analysis of gendered leadership and coupling in public schools. Accountability through Policy Multiple levels of the public education system are charged with producing and maintaining coupling structures, with much attention focused on the role of high-stakes accountability reforms handed down from various levels of government in recent years (Coburn 2004; Desimone 2013; Diamond 2007; Elmore, Abelman, and Furman 1996; Hallett 2010; Spillane and Burch 2006). In 1983, A Nation at Risk spurred the standards-based reform movement with claims of a “‘rising tide in mediocrity’ in U.S. Schools” (McDermott 2011, 60). The Clinton administration subsequently brought about the reauthorization of the Elementary Secondary Education Act (ESEA) as the Improving America’s Schools Act (IASA), and the Bush administration reauthorized ESEA/IASA as the No Child Left Behind Act of 2001 (NCLB). Unlike their predecessors, these policies redirected attention toward equity in education vis-à-vis performance accountability, and away from issues of desegregation, access, and equality in spending. Such federal policies and reports are fundamentally designed to affect all public schools. Geared toward performance accountability and improving academic achievement in public schools, policymakers assume that policy efforts will result in tangible changes or outcomes at the school level. However, many broad policies fail to take form within the walls of schools (Eagly and Johnson 1990; Pitner 1981; Shakeshaft 1987). While the federal policies from the mid-1980s through the first decade of the twenty-first century had variation in objectives, the overarching reforms centered on an increase in academic accountability standards for public schools across the nation, and are thus referred to as the standards-based reform movement (McDermott 2011). The federal policy reforms put forth by IASA and NCLB are highly focused on closing the achievement gap for racial minorities and poor students (Grissom, Kalogrides, and Loeb 2013), suggesting that the largest impact will be felt among schools that serve a greater proportion of minority and lower-class students. Organizational scholars recognize a changing educational landscape, and qualitative studies document a recoupling (Coburn 2004; Davies, Quirke, and Aurini 2006; Hallett 2010). Standards-based reforms have been specifically credited for changing how teachers deal with struggling students, teaching to the test, and classroom content (Desimone 2013). With few exceptions, contemporary educational research attributes the shifting trend in coupling structures (from looser to tighter), at least in part, to macro-level structures (e.g., federal policy, state accountability structures). Given the explicitly stated goals of federal policies, to increase academic accountability among public schools, I hypothesize that federal-level policies will tighten coupling between principals and teachers. State-level policies are also influential in shaping schools. State-level accountability structures can vary widely depending on the political climate of the state. Federal policies request test scores (e.g., NCLB’s Adequate Yearly Progress—AYP), but states have the power to increase standards, test more frequently, or standardize curriculums. States differ broadly along a continuum on which some states are low in high-stakes testing, others exhibit a great deal of high-stakes testing and accountability to the state, and still others fall somewhere in between (Carnoy and Loeb 2002). States retain a great deal of control over public schools and can dictate standards above and beyond the federal-level mandates. Though not as pervasive as federal-level policies, state accountability structures and laws often diffuse across states (for an example of charter school laws, see Renzulli and Roscigno [2005]). State governments will adopt similar policies (i.e., to their neighboring states), resulting in homogeneous regulations. While the intent of external pressures is to yield increased results in academic performance, it is possible that the degree of external pressures results in tighter coupling within the school, as teachers and administrators negotiate the requirements of the state. I hypothesize that state macro-level accountability structures play a role in tightening the micro-level coupling structures within schools. The Local Context Local governments and school boards oversee many organizational aspects within schools and still wield substantial control over their local schools (Diamond 2007), despite concerted centralization efforts by the federal and state governments. Macro structures at the federal and state level should trickle down to local schools, but local school boards and governments can require schools to adhere to additional district-level guidelines. For example, local governments may influence hiring decisions, curriculum content, or discipline policies. In these circumstances, the local district may play a central role in determining the internal structure of the school, and can serve as an intermediary between macro-level policies and changes within the school. Thus, various levels of the school system simultaneously affect day-to-day activities within schools, although they may affect different aspects of the school system. Local school districts often attempt to play an active role in shaping the internal structure of schools, and can penetrate the core activities inside classrooms (Coburn 2004; Hallett 2010). Such interplay between various hierarchical levels of the school system lend further credence to Orton and Weick’s (1990) proposition that we understand coupling in organizations through a dialectical approach. For instance, Hallett (2007b) found that the local school district will attempt to influence the organizational structure of a school by hiring a principal whose leadership style most closely aligns with their goals. Thus, the local tier of the school system may apply indirect pressures when placing administrators in schools. Neo-institutionalists would expect principals and teachers to decouple when faced with pressure from the local school district, and while some teachers do turn to decoupling strategies when faced with these institutional pressures, many others assimilate (among other strategies) (Coburn 2004). Moreover, the principal may work closely with the local district when the district is more integrated in the decision-making that affects the school. The local government is the meso-level of coupling, connecting the macro-level (federal and state) structures and the micro-level (school). While neo-institutionalists would predict that external pressure from the school district would result in looser coupling in schools, more recent studies on coupling suggest that district pressures may actually tighten coupling in the school (Diamond 2012). This may be especially true if principals actively push/support the school district’s agenda. Given the trends in recent literature, I hypothesize that tighter coupling from districts to schools will tighten the coupling between principals and teachers. The Role of the Principal: Female Leaders The institutional environment helps shape coupling in schools, but the principal plays a pivotal role in interpreting and enacting the mandates (Coburn 2004; Diamond 2007, 2012; Hallett 2010; Seyfarth and Bost 1982; Spillane et al. 2002; Young 2006). Principals are responsible for mediating policy messages, and this may disrupt teachers’ preferred structure within the school (Weiss and Cambone 1994). Even when the macro- or meso-level policies are touted as the overarching factor in shifting organizational structure of schools, the qualitative data illuminates the role of the principal (Diamond 2012; Spillane et al. 2002; Spillane, Parise, and Sherer 2011). Teachers commonly report losing autonomy (Diamond 2007; Hallett 2010) because principals do not let them “do their jobs” (Hallett 2010, 58). According to teachers, principals endorse and enforce policies within the school, exemplifying Orton and Weick’s (1990) dialectical understanding of coupling. While teachers report that principals do influence their classroom decisions, they also highlight how other forces are influential factors, such as other teachers, standards, testing, textbooks, and the Internet (Diamond 2007). Thus, principals do possess some agency to enact policy standards and changes, but if male and female principals embody different leadership styles, then recoupling or decoupling could also be a result of gendered leadership. A focus on the inner workings of schools—particularly principals—cannot ignore gender. Schools are gendered organizations; the gendered norms of schools disproportionately place males in administrator positions and females in teaching positions (Hallett 2007a; Lee, Smith, and Cioci 1993). Women have become principals in increasing numbers, making gendered leadership an increasingly critical issue in public schools. Principals display some of the same gendered management styles found in other organizational settings (Eagly and Karau 2002; Weiss and Cambone 1994). Female principals are more likely to communicate with teachers, stop into classrooms, walk through the hallways, and know the general pulse of the school (Fauth 1984; Ingersoll 1996; Pfeffer 1981; Shakeshaft 1987). Female principals engage in these behaviors due to their general interest in the teachers in their school. In contrast, males engage in traditional non-participatory management styles, relying on authoritative directives that are not followed up with communication, trips around the school, or time spent in classrooms (Lee, Smith, and Cioci 1993). In general, female principals wish to be “more prominently and personally involved,” compared to their male counterparts (Charters and Jovick 1981, 322). And while a democratic leadership style is often considered more effective (Eagly and Johnson 1990), the management of schools has shifted from a focus on educational philosophies toward that of bureaucracy and efficiency, an approach more consistent with other types of organizations, and may favor male principals. Based on accounts of leadership styles, I expect coupling in schools headed by female principals to be tighter than in those headed by males.1 Principals are in positions of power, and must acquire legitimacy for their leadership role while simultaneously maintaining organizational legitimacy for the school (Spillane, Hallett, and Diamond 2003). In some cases, this may only require assuming the title of “principal,” but this is unusual and principals tend to garner legitimacy when they can exhibit valued interaction styles (Spillane, Hallett, and Diamond 2003). When seeking legitimacy, female principals could turn to policies and regulations already viewed as legitimate in order to gain control over a school. Organizational actors can acquire legitimacy through sources of authority (Fauth 1984; Tyler 2006; Wingfield 2009; Zelditch 2001); for principals, authority may come from higher powers (e.g., policies, local government), and a principal could choose to embrace and implement policies in order to glean legitimacy (Spillane et al. 2002). Principals may also wish to send a signal to those individuals who oversee their work (district-level administrators). Districts want to hold schools accountable, but require a principal who is willing to enforce policies and maintain surveillance over the school (Hallett 2007b, 2010). Female principals may feel pressure to actively implement district policies in order to acquire legitimacy from those at the district level, but Hallett (2007b, 2010) demonstrates how principals who advocate for more tightly coupled relationships with teachers may face opposition and actually lose legitimacy with teachers. Context and circumstances may contribute to differences in gendered leadership because gendered leadership does not exist outside sociopolitical and local context. Multiple hierarchical levels can function harmoniously to shape the organizational coupling structure of schools; thus, I explore how female principals affect micro-coupling during federal policy eras with high accountability and in districts where tighter coupling exists between the district and the school. Data I draw on several data sources; primarily, the Schools and Staffing Survey (SASS), from the National Center for Education Statistics (NCES). This is the best dataset for a study of school trends in organizational structure and activities because it is a large, nationally representative sample of schools, principals, and teachers with measures of institutional environments and organizational structure. SASS includes seven waves of data spanning two and a half decades and multiple socio-political eras. The analyses shown in this manuscript draw on all seven waves of the SASS data (1987–1988, 1990–1991, 1993–1994, 1999–2000, 2003–2004, 2007–2008, 2011–2012). I refer to waves using the first part of the academic year (e.g., 1987–1988 is 1987). Due to changes in the questionnaire for the most recent wave of data (2011–2012), I am unable to include this wave in all of my analyses. I use this wave of data whenever possible, and in an appendix I include separate analyses that utilize this wave of data, but there are missing variables. When discussing findings, I also comment on how the last wave affects the results. The SASS data is a repeated cross-section of randomly sampled schools, resulting in a nationally representative sample of schools in each wave. For theoretical and empirical reasons, I exclude some schools and teachers from my analyses. After excluding all schools that are not traditional public schools (i.e., private schools, alternative schools, charter schools, or non-traditional schools), and all teachers in non-traditional classrooms, I am left with approximately 34,940 school-years (25,640 unique schools, with 7,040 repeatedly sampled) across six waves of data.2 For analyses that include the 2011 wave of data, I have approximately 40,360 school-years (28,910 unique schools, with 8,740 repeatedly sampled). Given the large number of cases, I also ran models using a random 10 percent subsample. Results of these models mirror those presented here. In a unique combination of data sources, I link the SASS with supplementary socio-political data that allow me to consider environmental factors that could influence schools. By including variables that address the external environment, I am better able to understand how the socio-political environment plays a role in shaping the organizational structuring of schools. Variables Dependent Variable Micro-Level Coupling (Loose-to-Tight Coupling) Micro-level coupling is a continuous variable, ranging from 0 to 4, where lower numbers indicate looser coupling and higher numbers indicate tighter coupling. Operationalizing a concept that is traditionally found in qualitative literature is a difficult task; thus, I focus on the elements highlighted in the scholarship on school coupling to create this variable. The primary signs of coupling can be found in elements of control and autonomy (i.e., how tightly linked administrative control is over what teachers do in their classrooms). This is consistent with previous research that studies the opposing concepts of connection and autonomy (Orton and Weick 1990), and recent educational research on coupling (Hallett 2010). This school-level coupling index is composed of five items indicating the extent of control/influence teachers have over: (1) choosing textbooks and other instructional materials, (2) choosing content, topics, and skills to be taught, (3) selecting teaching techniques, (4) disciplining students, and (5) determining the amount of homework assigned. These five items largely capture academic instruction, and are likely the places where change would occur if schools respond to the federal goals of performance accountability. Items are reported by teachers within each school in each of the waves. Teachers report how much individual influence/control they possess, and teachers responded to a Likert scale that ranged from “complete control/a great deal of control” (loosely coupled) to “no control” (tightly coupled). The SASS questionnaires asked identical questions across waves of data but offered different ordinal response categories to respondents in some waves of data. I standardized each item to have a 0–4 scale by multiplying it by the appropriate scaling factor (see table 1 for scaling factors) and averaged items to form a continuous micro-coupling index (α = 0.73). Schools are not just tightly or loosely coupled; they can be coupled on some dimensions and not others even in the same policy era. Therefore, I consider these five items as separate dependent variables. I do not show the models, but all models are substantively similar, ensuring that one or two items from the index are not driving the results presented here. Table 1. Scaling Factors for Standardizing Scales Original  Scaling factor  Rescaled  Waves: 1987, 1990, 1993  6-point scale  6 to 5  5-point  Range: 0–5    Range: 0–4  0  4/5  0  1  4/5  0.8  2  4/5  1.6  3  4/5  2.4  4  4/5  3.2  5  4/5  4  Original  Scaling factor  Rescaled  Waves: 2003, 2007, 2011  4 point Scale  4 to 5  5 point  Range: 0–3    Range: 0–4  0  1 1/3  0  1  1 1/3  1.34  2  1 1/3  2.67  3  1 1/3  4  Original  Scaling factor  Rescaled  Waves: 1987, 1990, 1993  6-point scale  6 to 5  5-point  Range: 0–5    Range: 0–4  0  4/5  0  1  4/5  0.8  2  4/5  1.6  3  4/5  2.4  4  4/5  3.2  5  4/5  4  Original  Scaling factor  Rescaled  Waves: 2003, 2007, 2011  4 point Scale  4 to 5  5 point  Range: 0–3    Range: 0–4  0  1 1/3  0  1  1 1/3  1.34  2  1 1/3  2.67  3  1 1/3  4  Note: All waves are scaled to the 1999 wave of SASS data. Teachers primarily consider their own classrooms, and the intra-class correlation among teachers (i.e., teachers who teach in the same school) is significant in all waves, ranging from 11 to 17 percent. This indicates the degree to which teachers agree on the organization of activities and autonomy within their school. To obtain a school-level measure of coupling, teacher reports of coupling are aggregated to the school level by averaging teachers’ reports within each school. On average, four to five teachers were interviewed from each school. The variation in teachers’ responses within the same school is less than the variation in teachers’ responses across schools; therefore, I am confident in the validity of using the aggregated responses of teachers to form a school-level variable. Federal-Level Independent Variables Policy Eras Time is an important dimension of this study, and I use the survey year as an indicator of policy era. I enter the variable as categorical to assess non-linearity in its effects, although for initial analyses on the trends of micro-coupling, I use a continuous policy-era variable that ranges from 0 to 6. Federal-level policies have changed over time. Survey year 1987 is the first wave of data available after the 1983 report A Nation at Risk, and the year 1987 is a reference category in all analyses. Survey year 1999 is the first wave of data after ESEA was reauthorized as IASA. Survey year 2003 is the first wave of data in the NCLB era, and while survey year 2007 still falls under the auspices of NCLB, much like 1990 and 1993 are still directly following A Nation at Risk, it is six years after the controversial policy, and schools are no longer transitioning into NCLB requirements. For those analyses that include the final wave of data, year 2011 is well after the enactment of NCLB, and by this point many states have sought waivers in place of NCLB requirements (US Department of Education 2013). SASS data are ideally situated for an analysis that considers the relationship between policy eras and micro-coupling. I expect time lags between policy implementation and micro-coupling, and if the waves of data are concurrent with policy changes, then the analyses may not truly capture the effects of the federal policy eras. NCLB passed in 2001, and was signed into law in early 2002; therefore, the 2003–2004 wave of SASS data is useful for capturing the early transitions into NCLB while still allowing for a lag between policy implementation and changes to micro-coupling. State-Level Independent Variables Index of High-Stakes Testing I am interested in accountability structures that states impose to improve test scores or control schools from the macro level. For accountability at the state level, I use Carnoy and Loeb’s (2002) state-accountability index. This index signifies the degree to which a state participates in high- or low-stakes testing. It “captures the degree of state external pressures on schools to improve student achievement according to state-defined performance criteria” (Carnoy and Loeb 2002, 311). This index ranges from 0 to 5, with 0 indicating a very low level of accountability within the state and 5 indicating a great deal of accountability. Though not time-varying, this is the best indicator available for accountability at the state level. Election Results by State I use the election maps provided by the US Electoral Maps (Office of the Federal Register 2013) to gauge the relative liberal or conservative leanings of each state. I coded the direction each state voted in the presidential election prior to the wave of data presented. (e.g., models with the 2007 wave of data use the 2004 presidential election). Each state is coded as 1 or 0, with 1 indicating that the state voted for the Republican candidate. (This is consistent with research that links state characteristics with educational data. See Renzulli and Roscigno [2005].) State Charter School Laws I use a dichotomous variable to indicate whether or not the state has a law permitting charter schools. Charter schools are not included in my analyses, but this information helps signal how states view educational choice movements. States adopted charter school laws in different years, and some never adopt charter school laws, making this a time-varying variable. Local-Level Independent Variables Meso-Level Coupling To capture meso-level coupling—coupling between the school district/school board and the school—I create an index consisting of three variables indicating the district’s control over hiring teachers in the school, setting curricula, and setting discipline policies for schools (α = 0.69). I recode this variable using the same format from my dependent variable. The categories range from “no influence” to “major influence,” and the coupling index between the local government and school is a loose-to-tight index, parallel to the micro-level (principal-to-teacher) coupling index (i.e., 0–4), but preliminary examination of this variable revealed non-normality. Preliminary models demonstrate non-linearity in the association between meso-level coupling and micro-coupling. Given this non-normality and evidence of non-linearity, I divide the meso-level coupling variable into quartiles, and enter this variable into the analyses as a categorical predictor. Quartile comparisons are then made via Wald tests (using alternative out-groups/reference categories). These tests reveal that the bottom three quartiles do not differ from one another in their association with micro-level coupling, but they each differ significantly from the fourth quartile in their associations. Hence, I account for this non-linearity by dividing my meso-level coupling variable into tight coupling (fourth quartile) or not (bottom three quartiles). Unfortunately, these measures are excluded in the most recent wave of data collection (2011), and models using this index are limited to data from waves 1987–2007. Merit Pay/Bonus I use a measure of merit pay to indicate whether or not the local district or school uses a system of performance-based merit pay or bonus to reward teachers for students’ test scores. This is a dichotomous variable (1 = merit pay/bonuses) and refers to the availability of a school-wide bonus for all teachers in a school with exceptional performance/improvement. Principal Characteristics Female Principal’s sex is a dichotomous variable (female = 1). Non-White Principal’s race is indicated with a dummy variable (nonwhite = 1) (White: 86 percent; Black: 7.8 percent; Hispanic: 3.3 percent; Asian: 1.3 percent; American Indian: 1.3 percent). Highest Degree Earned Most principals earn advanced degrees. I include the measure for doctorate degree in my models for highest degree earned, and all other degrees as the reference category. Principal Tenure The number of years a principal served ranges from 0 to 47. Principals who report 0 are serving their first year as principal, and the average is just over 5 years. Teaching Experience More than 99 percent of principals taught prior to becoming a principal. The length of time a principal spent teaching prior to becoming a principal ranges from 0 to 42 years. The average is 11.49 years, although females have a statistically significant higher average than males (13.59 vs. 10.49). Control Variables School-level and environmental factors could play a role in how school employees organize the internal activities of a school. The demographics of a school may motivate administrators, teachers, and staff to tighten or loosen coupling based on needs that are related or unrelated to policies, accountability structures, rewards, or the local school board. Controlling for these school-level factors reduces the likelihood that I am mis-specifying effects actually attributable to other organizational factors. I control for environmental characteristics, such as the location (i.e., northeast, midwest, south, and west), the setting (i.e., urban, suburban, and rural), grades served (i.e., elementary or secondary), enrollment, school demographics (i.e., percent free lunch, percent White, Black, Hispanic, Asian, and American Indian), and the percent of teachers who teach in tested subjects.3 Finally, due to the fact that some schools appear in multiple waves of the data, I also control for schools that are represented more than once over time. Table 2 provides additional information about each variable. Table 2. Variables of Interest: Definitions, Sources, and Descriptives (Means and Standard Deviations for All Seven Waves are Presented in Parentheses)  Variable  Description and Coding  Mean  SD  Dependent Variable  Micro-Level Coupling (Individual items are presented below)  Micro-level coupling is made up of 5 items that indicate the formal relationship between the principal and teachers. The following items are included in my scale: the extent of control/influence teachers have over (1) choosing textbooks and other instructional materials, (2) choosing content, topics, and skills to be taught, (3) selecting teaching techniques, (4) disciplining students, and (5) determining the amount of homework assigned. These items are reported by teachers within each school. Questions ask teachers to report how much individual influence or control they possess and teachers responded to a likert scale that ranged from “complete control/a great deal of control” (loosely coupled) to “no control” (tightly coupled). The scaled variable ranges from 0-4. α = 0.73.  0.91 (0.90)  0.47 (0.47)   1. Textbooks  Selecting textbooks and other instructional materials. Range:0-4. Higher numbers indicate tighter coupling.  1.42 (1.44)  0.87 (0.89)   2. Content  Selecting content, topics, and skills to be taught. Range:0-4. Higher numbers indicate tighter coupling.  1.34 (1.37)  0.85 (0.87)   3. Techniques  Selecting teaching techniques. Range:0-4. Higher numbers indicate tighter coupling.  0.50 (0.51)  0.50 (0.51)   4. Discipline  Disciplining students. Range:0-4. Higher numbers indicate tighter coupling.  0.84 (0.83)  0.58 (0.58)   5. Homework  Determining the amount of homework to be assigned. Range:0-4. Higher numbers indicate tighter coupling.  0.43 (0.43)  0.50 (0.50)  Independent Variables  Federal Level      Survey Waves  Continuous variable ranges from 0 to 6, and represents each wave in the SASS data (1987, 1990, 1993, 1999, 2003, 2007, and 2011).  2.48 (2.94)  1.65 (1.94)  Year 1987  1987-1988 survey from the SASS questionnaires. (A Nation at Riskera) (Reference category).  0.15 (0.13)  0.36 (0.33)  Year 1990  1990-1991 survey from the SASS questionnaires.  0.18 (0.16)  0.39 (0.37)  Year 1993  1993-1994 survey from the SASS questionnaire (IASA era).  0.18 (0.15)  0.38 (0.36)  Year 1999  1999-2000 survey from the SASS questionnaire.  0.17 (0.15)  0.38 (0.36)  Year 2003  2003-2004 survey from the SASS questionnaire (NCLB era).  0.17 (0.13)  0.38 (0.36)  Year 2007  2007-2008 survey from the SASS questionnaire.  0.15 (0.13)  0.35 (0.33)  Year 2011  2011-2012 survey from the SASS questionnaire.  ---- (0.13)  ---- (0.34)    State Level      Index of High Stakes Testing  This index ranges from 0-5, and represents the extent each state has external accountability. 0 indicates a low level of accountability, and 5 is the highest level of accountability. Source: (Carnoy and Loeb 2002).  2.25 (2.26)  1.49 (1.49)  Election Results  Each state is coded as 1 for republican or 0 for democrat based on how they voted in the most recent presidential election. Data from the 2007 wave uses election results from the 2004 election.  0.64 (0.61)  0.48 (0.49)  State Charter School Law  This is a dichotomous variable that indicates whether or not a state has a law that allows charter schools. States adopt charter school laws at different times; thus, this variable changes over time. States with charter school laws are coded as 1 once they acquire the law.  0.41 (0.47)  0.49 (0.50)    District Level      Meso-Level Coupling  This level of coupling is the relationship between the district and the school. This variable is a scale made up of 3 items that range from 0-4, but due to non-normality, I include a categorical variable that indicates if the meso-level coupling is in the highest quartile (very tight) and is coded as 1. α = 0.61.  0.21 (---)  0.40 (---)  Bonus  This is a measure of merit pay that indicates whether or not the local district or school uses a system of performance based merit pay or a bonus to reward teachers for their students’ test scores. This is a dichotomous variable and refers to the availability of a school wide bonus for all teachers in a school with exceptional performance or improvement. The measure of school wide bonuses also considers incentives outside of the school that are present in the district.  0.09 (0.09)  0.29 (0.28)    Principal Level      Gender  Female principals are coded as 1.  0.31 (0.32)  0.46 (0.47)  Race  Nonwhite principals are coded as 1. White principals are coded as 0. Non-white principals only make up 13% of the data.  0.13 (0.13)  0.34 (0.34)  Highest Degree Earned  The highest degree the principal has earned. Degrees range from Bachelors to Masters, Specialist, and Doctorates. I code doctorate degrees as 1, and use all other degrees as the reference category.  0.09 (0.09)  0.29 (0.29)  Number of Years as Principal  The number of years a principal has served as a principal.  5.13 (5.04)  5.40 (5.30)  Number of Years Teaching Prior to Principal  The number of years a principal spent teaching in the classroom prior to becoming a principal.  11.45 (11.49)  6.36 (6.37)  Control Variables  Northeast  School is located in the northeast.  0.16 (0.16)  0.37 (0.37)  Midwest  School is located in the midwest.  0.26 (0.26)  0.44 (0.44)  West  School is located in the west.  0.25 (0.25)  0.43 (0.43)  South  School is located in the south (Reference category).  0.34 (0.33)  0.47 (0.47)  Urban  School is located in an urban setting.  0.20 (0.20)  0.40 (0.40)  Suburban  School is located in a suburban setting.  0.31 (0.31)  0.46 (0.46)  Rural  School is located in a rural setting (Reference category).  0.49 (0.50)  0.50 (0.50)  Grades Served  Grades served in the school. Elementary schools are coded as 1. Combined and secondary schools are coded as 0.  0.55 (0.55)  0.50 (0.50)  Enrollment  The number of students enrolled in the school.  633.67 (642.86)  497.18 (504.32)  Percent Free Lunch  This variable is the percentage of students who are eligible to receive free lunch in the school (range: 0-100).  36.07 (37.16)  26.10 (26.41)  Percent White  Percent of White students enrolled in the school. (Reference category).  73.17 (72.47)  29.99 (29.99)  Percent Black  Percent of Black students enrolled in the school.  12.05 (11.91)  22.07 (21.73)  Percent Hispanic  Percent of Hispanic students enrolled in the school.  8.35 (9.05)  17.46 (18.05)  Percent Asian  Percent of Asian students enrolled in the school.  2.78 (2.79)  9.30 (8.93)  Percent American Indian  Percent of American Indian students enrolled in the school.  3.65 (3.52)  13.54 (13.23)  Teachers in a Tested Subject  The percent of teachers who responded within the school, and teach in a tested subject.  48.54 (49.04)  33.33 (33.17)  Duplicate School  Some schools apper more than once across all waves of the SASS data. I control for schools that appear more than once (coded as 1).  0.49 (0.49)  0.50 (0.50)  N = 34,940 (N=40,360)  (Means and Standard Deviations for All Seven Waves are Presented in Parentheses)  Variable  Description and Coding  Mean  SD  Dependent Variable  Micro-Level Coupling (Individual items are presented below)  Micro-level coupling is made up of 5 items that indicate the formal relationship between the principal and teachers. The following items are included in my scale: the extent of control/influence teachers have over (1) choosing textbooks and other instructional materials, (2) choosing content, topics, and skills to be taught, (3) selecting teaching techniques, (4) disciplining students, and (5) determining the amount of homework assigned. These items are reported by teachers within each school. Questions ask teachers to report how much individual influence or control they possess and teachers responded to a likert scale that ranged from “complete control/a great deal of control” (loosely coupled) to “no control” (tightly coupled). The scaled variable ranges from 0-4. α = 0.73.  0.91 (0.90)  0.47 (0.47)   1. Textbooks  Selecting textbooks and other instructional materials. Range:0-4. Higher numbers indicate tighter coupling.  1.42 (1.44)  0.87 (0.89)   2. Content  Selecting content, topics, and skills to be taught. Range:0-4. Higher numbers indicate tighter coupling.  1.34 (1.37)  0.85 (0.87)   3. Techniques  Selecting teaching techniques. Range:0-4. Higher numbers indicate tighter coupling.  0.50 (0.51)  0.50 (0.51)   4. Discipline  Disciplining students. Range:0-4. Higher numbers indicate tighter coupling.  0.84 (0.83)  0.58 (0.58)   5. Homework  Determining the amount of homework to be assigned. Range:0-4. Higher numbers indicate tighter coupling.  0.43 (0.43)  0.50 (0.50)  Independent Variables  Federal Level      Survey Waves  Continuous variable ranges from 0 to 6, and represents each wave in the SASS data (1987, 1990, 1993, 1999, 2003, 2007, and 2011).  2.48 (2.94)  1.65 (1.94)  Year 1987  1987-1988 survey from the SASS questionnaires. (A Nation at Riskera) (Reference category).  0.15 (0.13)  0.36 (0.33)  Year 1990  1990-1991 survey from the SASS questionnaires.  0.18 (0.16)  0.39 (0.37)  Year 1993  1993-1994 survey from the SASS questionnaire (IASA era).  0.18 (0.15)  0.38 (0.36)  Year 1999  1999-2000 survey from the SASS questionnaire.  0.17 (0.15)  0.38 (0.36)  Year 2003  2003-2004 survey from the SASS questionnaire (NCLB era).  0.17 (0.13)  0.38 (0.36)  Year 2007  2007-2008 survey from the SASS questionnaire.  0.15 (0.13)  0.35 (0.33)  Year 2011  2011-2012 survey from the SASS questionnaire.  ---- (0.13)  ---- (0.34)    State Level      Index of High Stakes Testing  This index ranges from 0-5, and represents the extent each state has external accountability. 0 indicates a low level of accountability, and 5 is the highest level of accountability. Source: (Carnoy and Loeb 2002).  2.25 (2.26)  1.49 (1.49)  Election Results  Each state is coded as 1 for republican or 0 for democrat based on how they voted in the most recent presidential election. Data from the 2007 wave uses election results from the 2004 election.  0.64 (0.61)  0.48 (0.49)  State Charter School Law  This is a dichotomous variable that indicates whether or not a state has a law that allows charter schools. States adopt charter school laws at different times; thus, this variable changes over time. States with charter school laws are coded as 1 once they acquire the law.  0.41 (0.47)  0.49 (0.50)    District Level      Meso-Level Coupling  This level of coupling is the relationship between the district and the school. This variable is a scale made up of 3 items that range from 0-4, but due to non-normality, I include a categorical variable that indicates if the meso-level coupling is in the highest quartile (very tight) and is coded as 1. α = 0.61.  0.21 (---)  0.40 (---)  Bonus  This is a measure of merit pay that indicates whether or not the local district or school uses a system of performance based merit pay or a bonus to reward teachers for their students’ test scores. This is a dichotomous variable and refers to the availability of a school wide bonus for all teachers in a school with exceptional performance or improvement. The measure of school wide bonuses also considers incentives outside of the school that are present in the district.  0.09 (0.09)  0.29 (0.28)    Principal Level      Gender  Female principals are coded as 1.  0.31 (0.32)  0.46 (0.47)  Race  Nonwhite principals are coded as 1. White principals are coded as 0. Non-white principals only make up 13% of the data.  0.13 (0.13)  0.34 (0.34)  Highest Degree Earned  The highest degree the principal has earned. Degrees range from Bachelors to Masters, Specialist, and Doctorates. I code doctorate degrees as 1, and use all other degrees as the reference category.  0.09 (0.09)  0.29 (0.29)  Number of Years as Principal  The number of years a principal has served as a principal.  5.13 (5.04)  5.40 (5.30)  Number of Years Teaching Prior to Principal  The number of years a principal spent teaching in the classroom prior to becoming a principal.  11.45 (11.49)  6.36 (6.37)  Control Variables  Northeast  School is located in the northeast.  0.16 (0.16)  0.37 (0.37)  Midwest  School is located in the midwest.  0.26 (0.26)  0.44 (0.44)  West  School is located in the west.  0.25 (0.25)  0.43 (0.43)  South  School is located in the south (Reference category).  0.34 (0.33)  0.47 (0.47)  Urban  School is located in an urban setting.  0.20 (0.20)  0.40 (0.40)  Suburban  School is located in a suburban setting.  0.31 (0.31)  0.46 (0.46)  Rural  School is located in a rural setting (Reference category).  0.49 (0.50)  0.50 (0.50)  Grades Served  Grades served in the school. Elementary schools are coded as 1. Combined and secondary schools are coded as 0.  0.55 (0.55)  0.50 (0.50)  Enrollment  The number of students enrolled in the school.  633.67 (642.86)  497.18 (504.32)  Percent Free Lunch  This variable is the percentage of students who are eligible to receive free lunch in the school (range: 0-100).  36.07 (37.16)  26.10 (26.41)  Percent White  Percent of White students enrolled in the school. (Reference category).  73.17 (72.47)  29.99 (29.99)  Percent Black  Percent of Black students enrolled in the school.  12.05 (11.91)  22.07 (21.73)  Percent Hispanic  Percent of Hispanic students enrolled in the school.  8.35 (9.05)  17.46 (18.05)  Percent Asian  Percent of Asian students enrolled in the school.  2.78 (2.79)  9.30 (8.93)  Percent American Indian  Percent of American Indian students enrolled in the school.  3.65 (3.52)  13.54 (13.23)  Teachers in a Tested Subject  The percent of teachers who responded within the school, and teach in a tested subject.  48.54 (49.04)  33.33 (33.17)  Duplicate School  Some schools apper more than once across all waves of the SASS data. I control for schools that appear more than once (coded as 1).  0.49 (0.49)  0.50 (0.50)  N = 34,940 (N=40,360)  Note: Data is from the National Center for Education Statistics (NCES) Schools and Staffing Survey (SASS). The sample uses all traditional schools in the 1987, 1990, 1993, 1999, 2003, 2007, and 2011 waves of SASS data that have valid information on my independent, dependent, and control variables. I exclude private, charter, alternative, and non-traditional schools. Analytic Strategy This analysis proceeds in two major steps. First, using all seven waves of data from 1987 to 2011, table 3 examines the general trends in micro-coupling over time. In model 1, I use a continuous measure of time to examine the overall effect of time on micro-level coupling. In model 2, I break the survey years into separate dummy variables in order to evaluate how each time period affects micro-level coupling. Second, using the pooled waves from 1987 to 2007 and clustered robust standard errors to account for the nesting of schools within states (ICC = 0.10), I use OLS regression to analyze how the tiered levels of the educational system contribute to micro-level coupling. I utilize a stepwise approach entering each of the tiered levels of interest—federal eras and state laws/accountability structures, local characteristics, and principal attributes—in a sequential fashion. Table 3. OLS Regression of Micro-Coupling on Federal Policy Eras   Model 1  Model 2  Socio-political factors      Survey waves (1987–2011)  0.007**    Year 1990    −0.050***  Year 1993    −0.065***  Year 1999    0.041**  Year 2003    −0.076***  Year 2007    −0.032*  Year 2011    0.049***  (Reference is 1987 (ANaR))      Controls      Location      Northeast  −0.099*  −0.104*  Midwest  −0.172***  −0.175***  West  −0.116**  −0.116**  (reference is South)      Setting      Urban  0.175***  0.179***  Suburban  0.124***  0.129***  (reference is Rural)      School demographics      Free lunch percentage  0.000  0.000  Black percentage  0.003***  0.003***  Hispanic percentage  0.002***  0.002***  Asian percentage  −0.002*  −0.002*  American Indian percentage  0.001  0.001  (reference is White percentage)      Enrollment  0.000***  0.000***  Elementary school  0.208***  0.215***  Percent of teachers in tested subject  0.002***  0.002***  Duplicate school  −0.019**  −0.011  (reference is Non-repeated school)      Constant  0.579***  0.631***  N = 40,360      R-squared  0.215  0.223    Model 1  Model 2  Socio-political factors      Survey waves (1987–2011)  0.007**    Year 1990    −0.050***  Year 1993    −0.065***  Year 1999    0.041**  Year 2003    −0.076***  Year 2007    −0.032*  Year 2011    0.049***  (Reference is 1987 (ANaR))      Controls      Location      Northeast  −0.099*  −0.104*  Midwest  −0.172***  −0.175***  West  −0.116**  −0.116**  (reference is South)      Setting      Urban  0.175***  0.179***  Suburban  0.124***  0.129***  (reference is Rural)      School demographics      Free lunch percentage  0.000  0.000  Black percentage  0.003***  0.003***  Hispanic percentage  0.002***  0.002***  Asian percentage  −0.002*  −0.002*  American Indian percentage  0.001  0.001  (reference is White percentage)      Enrollment  0.000***  0.000***  Elementary school  0.208***  0.215***  Percent of teachers in tested subject  0.002***  0.002***  Duplicate school  −0.019**  −0.011  (reference is Non-repeated school)      Constant  0.579***  0.631***  N = 40,360      R-squared  0.215  0.223  Note: Schools are clustered in states. Data are from the National Center for Education Statistics (NCES) Schools and Staffing Survey (SASS) and, following NCES convention, I have rounded sample size numbers to the nearest 10 in order to protect the identities of respondents. Given the large number of cases, models were examined using a random 10 percent subsample, and the results are substantively similar. The sample uses all traditional schools in the 1987, 1990, 1993, 1999, 2003, 2007, and 2011 waves of SASS data that have valid information on my independent, dependent, and control variables. I exclude private, charter, alternative, and non-traditional schools. *p < 0.05 **p < 0.01 ***p < 0.001 Due to data limitations, I am unable to present table 4 using data from the 2011 wave of data. In an  appendix, I include the 2011 survey year in models that largely replicate those presented in table 4. The meso-level coupling variable and the interaction term for gender and meso-level coupling are omitted from these models. Table 4. OLS Regression of Micro-Coupling on Federal, State, Local, and Principal Characteristics   Model 1  Model 2  Model 3  Model 4  Model 5  Socio-political factors            Year 1990  −0.052***  −0.050***  −0.054***  −0.048***  −0.054***  Year 1993  −0.075***  −0.070***  −0.077***  −0.058***  −0.077***  Year 1999 (IASA)  0.038*  0.042*  0.031  0.040*  0.031  Year 2003 (NCLB)  −0.076***  −0.077***  −0.092***  −0.118***  −0.092***  Year 2007  −0.026  −0.027  −0.042*  −0.073***  −0.042*  Reference is 1987 (ANaR)            State characteristics            Index of high-stakes testing  0.013  0.013  0.012  0.013  0.012  Republican state (reference is Democrat)  −0.022  −0.021  −0.020  −0.021  −0.020  Charter law  −0.005  −0.006  −0.007  −0.007  −0.007  Local characteristics            Tight meso-coupling (local govt to school relationship)    0.033***  0.031***  0.031***  0.018*  (Reference is 0–3)            Bonus    −0.012  −0.012  −0.005  −0.012  Principal’s characteristics            Female principal (reference is Male)      0.060***  0.055**  0.052***  Non-White principal (reference is White)      −0.003  −0.001  −0.004  Highest degree—Doctorate      0.003  0.002  0.003  (reference is All other degrees)            Principal tenure (in years)      −0.002***  −0.002***  −0.002***  Teaching experience (in years) prior to principal      0.000  0.000  0.000  Interaction terms            Female principal x Year 1990        −0.029    Female principal x Year 1993        −0.079***    Female principal x Year 1999        −0.027    Female principal x Year 2003        0.074**    Female principal x Year 2007        0.074**    Female principal x Tightest meso-coupling          0.036*  Controls            Location            Northeast  −0.099*  −0.099*  −0.098*  −0.098*  −0.098*  Midwest  −0.155**  −0.155**  −0.152**  −0.150**  −0.153**  West  −0.094  −0.093  −0.097  −0.095  −0.097  (reference is South)            Setting            Urban  0.184***  0.182***  0.175***  0.176***  0.175***  Suburban  0.128***  0.127***  0.122***  0.123***  0.122***  (reference is Rural)            School demographics            Free lunch percentage  −0.000  −0.000  −0.000  −0.000  −0.000  Black percentage  0.003***  0.003***  0.003***  0.003***  0.003***  Hispanic percentage  0.001***  0.001***  0.001**  0.001**  0.001**  Asian percentage  −0.002**  −0.002**  −0.002**  −0.002**  −0.002**  American Indian percentage  0.001  0.001  0.001  0.001  0.001  (reference is White percentage)            Enrollment  0.000***  0.000***  0.000***  0.000***  0.000***  Elementary school (reference is Secondary/Combined schools)  0.211***  0.211***  0.199***  0.198***  0.199***  Percent of teachers in tested subject  0.002***  0.002***  0.002***  0.002***  0.002***  Duplicate school  −0.002  −0.007  −0.005  0.000  −0.005  (reference is Non-repeated school)            Constant  0.628***  0.624***  0.636***  0.630***  0.638***  N = 34,940            R-squared  0.221  0.221  0.225  0.228  0.226    Model 1  Model 2  Model 3  Model 4  Model 5  Socio-political factors            Year 1990  −0.052***  −0.050***  −0.054***  −0.048***  −0.054***  Year 1993  −0.075***  −0.070***  −0.077***  −0.058***  −0.077***  Year 1999 (IASA)  0.038*  0.042*  0.031  0.040*  0.031  Year 2003 (NCLB)  −0.076***  −0.077***  −0.092***  −0.118***  −0.092***  Year 2007  −0.026  −0.027  −0.042*  −0.073***  −0.042*  Reference is 1987 (ANaR)            State characteristics            Index of high-stakes testing  0.013  0.013  0.012  0.013  0.012  Republican state (reference is Democrat)  −0.022  −0.021  −0.020  −0.021  −0.020  Charter law  −0.005  −0.006  −0.007  −0.007  −0.007  Local characteristics            Tight meso-coupling (local govt to school relationship)    0.033***  0.031***  0.031***  0.018*  (Reference is 0–3)            Bonus    −0.012  −0.012  −0.005  −0.012  Principal’s characteristics            Female principal (reference is Male)      0.060***  0.055**  0.052***  Non-White principal (reference is White)      −0.003  −0.001  −0.004  Highest degree—Doctorate      0.003  0.002  0.003  (reference is All other degrees)            Principal tenure (in years)      −0.002***  −0.002***  −0.002***  Teaching experience (in years) prior to principal      0.000  0.000  0.000  Interaction terms            Female principal x Year 1990        −0.029    Female principal x Year 1993        −0.079***    Female principal x Year 1999        −0.027    Female principal x Year 2003        0.074**    Female principal x Year 2007        0.074**    Female principal x Tightest meso-coupling          0.036*  Controls            Location            Northeast  −0.099*  −0.099*  −0.098*  −0.098*  −0.098*  Midwest  −0.155**  −0.155**  −0.152**  −0.150**  −0.153**  West  −0.094  −0.093  −0.097  −0.095  −0.097  (reference is South)            Setting            Urban  0.184***  0.182***  0.175***  0.176***  0.175***  Suburban  0.128***  0.127***  0.122***  0.123***  0.122***  (reference is Rural)            School demographics            Free lunch percentage  −0.000  −0.000  −0.000  −0.000  −0.000  Black percentage  0.003***  0.003***  0.003***  0.003***  0.003***  Hispanic percentage  0.001***  0.001***  0.001**  0.001**  0.001**  Asian percentage  −0.002**  −0.002**  −0.002**  −0.002**  −0.002**  American Indian percentage  0.001  0.001  0.001  0.001  0.001  (reference is White percentage)            Enrollment  0.000***  0.000***  0.000***  0.000***  0.000***  Elementary school (reference is Secondary/Combined schools)  0.211***  0.211***  0.199***  0.198***  0.199***  Percent of teachers in tested subject  0.002***  0.002***  0.002***  0.002***  0.002***  Duplicate school  −0.002  −0.007  −0.005  0.000  −0.005  (reference is Non-repeated school)            Constant  0.628***  0.624***  0.636***  0.630***  0.638***  N = 34,940            R-squared  0.221  0.221  0.225  0.228  0.226  Note: Schools are clustered in states. Data are from the National Center for Education Statistics (NCES) Schools and Staffing Survey (SASS) and, following NCES convention, I have rounded sample size numbers to the nearest 10 in order to protect the identities of respondents. Given the large number of cases, models were examined using a random 10 percent subsample, and the results are substantively similar. The sample uses all traditional schools in the 1987, 1990, 1993, 1999, 2003, and 2007 waves of SASS data that have valid information on my independent, dependent, and control variables. I exclude private, charter, alternative, and non-traditional schools. *p < 0.05 **p < 0.01 ***p < 0.001 Results The analyses presented in table 3 test the direct associations between federal policy era and micro-level coupling. Model 1 indicates a significant and positive relationship between time and micro-level coupling, suggesting that schools have tightened up coupling throughout the standards-based reform era, generally supporting previous findings (Hallett 2010). In model 2 of table 3, I include each survey wave as a dummy variable in order to explore the nuance of how the standards-based reform era affected micro-level coupling. This model shows that schools have not simply tightened their coupling over time; instead, the relationship is non-linear. I find a significant positive association between the 1999 and 2011 waves and micro-level coupling. Relative to 1987, the 1999 and 2011 waves are the only two eras with a positive association to micro-level coupling, suggesting periods of recoupling. The NCLB era (2003), relative to 1987, has a significant negative association with micro-coupling, and that association persists for the 2007 era, indicating a period of decoupling. Models 1 and 2 in table 3 demonstrate an interesting pattern across US public schools, and speak to the general trends in coupling. Current developments in educational research suggest a recoupling movement in public schools, and while model 1 depicts a small but steady increase in coupling across public schools, model 2 tells a more nuanced story about the relative effects of policy eras and illustrates a non-linear trend with schools experiencing decoupling, recoupling, decoupling, and finally recoupling again. Relevant federal reports and policies (i.e., A Nation at Risk, IASA, and NCLB) are noted in figure 1 to provide reference points. The trends in coupling help place the recent qualitative studies of recoupling and increased tighter coupling into context. Wald tests for model 2 in table 3 reveal that most contrasts between successive years are significantly different from one another (with 1990 to 1993 as exceptions). Therefore, the observed recoupling from 2003 to 2007 is a statistically significant change. The recoupling that is observed in 2007 does suggest that NCLB may have had longer-term effects on the organization of public schools. The recoupling trend continues from 2007 through 2011, and this is well past the implementation of NCLB. Figure 1. View largeDownload slide Graph of micro-coupling over time Note: Estimates based on model 2 in table 3. Figure 1. View largeDownload slide Graph of micro-coupling over time Note: Estimates based on model 2 in table 3. In table 4, I include the remaining tiers of the public education system. Model 1 of table 4 introduces state-level characteristics. None of the state characteristics are significantly associated with micro-level coupling. Recall that these school-level data are clustered by state and the statistical significance of the ICC denotes the importance of state-level characteristics. While state-level features modeled in table 4 are non-significant, 10 percent of the variance between schools is due to state-level differences not captured here. Model 1 of  table A1 (see the appendix) replicates this analysis and includes the 2011 wave of data. Model 2 steps in the meso-level structures, which include local-to-school coupling relationships and the presence of a performance-based merit pay or bonus system. As expected, tight meso-level coupling is positively associated with tighter micro-level coupling. Model 2 suggests that when meso-level coupling is tightest (i.e., in the fourth quartile), there is an increase of 0.033 on micro-level coupling. The micro-level coupling index ranges from 0 to 4, with a mean of 0.90, making 0.033 (approximately 7 percent of a standard deviation) a moderate increase in micro-level coupling. Most federal policy eras are significant in model 2 (2007 is the exception), suggesting that both federal and local elements shape school environments. Model 2 of  table A1 reproduces these models using the 2011 data (and excludes meso-level coupling). The results in  table A1 are substantively similar, and the survey year of 2011 is positively associated with tighter micro-level coupling. Model 3 brings together all levels of the public school system and explores the role of the principal. Schools with female principals are more tightly coupled at the micro-level. Although many other principal attributes are non-significant, the variable indicating principals’ tenure is negatively related to tight coupling. For every additional year a principal has served as a principal, there is a decrease of 0.002 in micro-coupling. Meso-level coupling remains positively significant across models. Most federal policy eras are also still significant in model 3 (except 1999); model 3 in  table A1 replicates these models using 2011 data (and excludes meso-level coupling) and shows a similar pattern, although the survey year of 2011 is not significant in this model. While model 3 conveys how all levels of socio-political factors directly affect schools, it does not adequately speak to the potential interaction between these levels of analysis. I ran two sets of interactions to explore how contexts matter for gendered effects. First, I ran an interaction between federal policy eras and female principals (model 4). Second, I ran an interaction between meso-level coupling and female principals (model 5). The interactions in model 4 of table 4 illustrate how federal policy eras vary by gendered leadership. Male and female principals are similar in many policy eras, but their management styles do vary in some contexts. Specifically, the interaction terms for the most recent waves (2003 and 2007) are significant and positive. Compared to their male counterparts, female principals in 2003 and 2007 are associated with tighter micro-level coupling. In model 4 of  table A1, I also model interactions that include the 2011 wave of data. The results are similar, but the interaction term for female principals in 2011 is non-significant. Figure 2 graphs the patterns of these interactions, and Wald tests reveal that most interaction terms are significantly different from one another (2003 to 2007 is the exception). I use a dotted line to graph the 2011 data because it is based on the analyses in model 4 of  table A1. Figure 2. View largeDownload slide Graph of interaction between principal’s gender and federal policy eras Note: The solid line from 1987 to 2007 uses data from model 4 in table 4. The dotted line from 2007 to 2011 uses data from model 4 in table A1 (see appendix). Figure 2. View largeDownload slide Graph of interaction between principal’s gender and federal policy eras Note: The solid line from 1987 to 2007 uses data from model 4 in table 4. The dotted line from 2007 to 2011 uses data from model 4 in table A1 (see appendix). Model 5 includes the interaction for meso-level coupling and gender, and the significant coefficient for the interaction means the association between meso-level coupling and micro-level coupling does vary by principals’ gender. Tight meso-level coupling is associated with a 0.054 point increase in micro-level coupling when the principal is female (bmeso + bmesoxfemale = 0.018 + 0.036 = 0.054), and only a 0.018 point increase in micro-level coupling when the principal is male. Figure 3 shows that compared to male principals, female principals are associated with tighter micro-level coupling directly, and female principals also strengthen the relationship between local-level coupling and micro-level coupling. A slope test revealed that the relationship between meso-level coupling and micro-level coupling is statistically significant for both male and female principals, but significantly stronger for female principals. I explored interactions between female principals and other school factors (i.e., race, urbanicity), but these were not significant and are thus not included here. Figure 3. View largeDownload slide Graph of interaction between principal’s gender and meso-level coupling Figure 3. View largeDownload slide Graph of interaction between principal’s gender and meso-level coupling The analyses reveal fascinating findings pertaining to school demographic characteristics. First, the general socioeconomic status, measured through free lunch eligibility, is not a significant predictor of tightly coupled schools, but the racial composition of the school is significantly associated with tighter coupling. As the percentage of Black and Hispanic students increases within the school, schools experience tighter coupling. Second, elementary schools (compared to secondary or combined schools) are positively associated with tighter coupling. Third, school setting is associated with coupling, and urban schools have tighter coupling compared to rural and suburban schools. Fourth, as the percentage of responding teachers who teach in tested subjects increases, we see tighter micro-level coupling. Finally, schools that appear multiple times in my data are not significantly affecting the results presented here. When I include the 2011 wave of data ( table A1 in the appendix), the findings are reproduced. These results tell a thought-provoking story about how multiple levels of the education system intersect to influence school coupling. Discussion and Conclusion These results challenge some of our taken-for-granted assumptions about school coupling and highlight the necessity of integrating literatures to provide a more complete picture. Empirically, many studies of coupling focus on the outcomes of tight coupling in the context of shifting policy environments. This study uses those analyses to provide a framework and a backdrop for evaluating the impetus for tighter coupling on a national level. I measure how different levels of the public education system—including gendered leadership structures within schools—contribute independently and interactively to tighter coupling within public schools. Tighter coupling is not merely a result of one hierarchical level exerting an overwhelming force on the interior structure of schools; rather, multiple levels must be considered simultaneously. These results suggest that a number of federal-level policy eras are more relevant than others. State-level factors had no significant relationship with school-level organization, despite the overwhelming credit often bestowed upon states and “states’ rights” in education policy. Local government coupling relationships (i.e., meso-level coupling) influence tighter coupling at the micro-level; stronger ties at the local level encourage tighter coupling within the school. Many principal characteristics are not significantly related to tightly coupled structures within schools. Principals’ tenure and gender both directly predict coupling. In fact, schools with female principals have tighter coupling than those with male principals. Finally, important intersections between context and gendered leadership help detail the comprehensive story of school coupling. This research uniquely addressed two competing ideas from organizational research. First, coupling scholars have found that creating or increasing accountability standards has engendered tighter coupling within schools. Second, neo-institutionalism posits that chaos and disorder will stimulate an environment of loose coupling within an organization. The standards-based reform era, specifically for the policy era of NCLB, provides an interesting situation where increased accountability and chaos occurred in the same time period. Federal policies structured around performance accountability gradually increased demands on public schools, culminating in the NCLB era. Unfortunately, NCLB is considered poorly executed through direction, funding, and mandates (Mathis 2003; Orlich 2004; Weeden 2005). Scholars criticize NCLB as chaotic and confusing for schools, principals, and teachers (Cochran-Smith and Lytle 2006; Darling-Hammond 2007a, 2007b; Le Floch, Taylor, and Thomsen 2006; Valli and Buese 2007). Organizational scholars discuss the implications of chaotic environments and predict that organizational actors will actively loosen structures of coupling when faced with chaotic institutional environments (Weick 1976). The presence of a federal policy that simultaneously produces increased accountability standards and chaos for schools creates a tension for this research and organizational research in general. My findings suggest that the transition into NCLB is negatively associated with tight coupling, and the confusion surrounding the federal policy was a stronger force than the pressure of accountability standards. This result is consequential, but it is possible that it is distinct to public schools. Public schools could be an exceptional organizational form, and not all organizations may respond similarly when faced with competing pressures of accountability and chaos. The general trend in micro-level coupling is non-linear, and while organizational scholars know organizations can change (Aldrich and Reuf 2006), it is unclear how often an internal organizational structure shifts. With regard to my findings on US public schools, coupling within schools has changed over time, but it has not simply steadily increased (or decreased) over time. The picture presented in this paper suggests movement, but it is also possible that change is temporary or never strays too far from an average. How changes in micro-level coupling affect daily life in schools is difficult to discern from these analyses, but are well supplemented by the qualitative research conducted during these policy eras. Considering this research in light of those studies helps show patterns in coupling, while simultaneously allowing us to reflect on the processes behind the changes in coupling. My analyses do not show large effects in any one time period. Indeed, all of the coefficients are smaller than 0.12, although across multiple time periods (e.g., from 1993 to 1999) the change in micro-level coupling is greater than 0.10. With a standard deviation of 0.47 for micro-coupling, a change slightly greater than 0.10 indicates just under one-quarter of a standard deviation change from 1993 to 1999. The exact meaning behind one-fifth to one-quarter of a standard deviation is not evidenced in the models presented here, but several studies were conducted during the highest peak illustrated in figure 1 (see, for instance, Coburn 2004; Diamond 2007; Hallett 2010), and this research helps exhibit what tighter coupling looks like in teachers’ day-to-day lives. These studies help put my nationally representative findings into context, and imply that even what appears to be a small to moderate shift from looser to tighter coupling within schools does effect change within their walls. If policymakers hope principals and teachers are forming tight linkages within schools, then the non-linear relationship between federal policy eras and micro-coupling presents a unique dilemma for shaping future federal-level policies. There was a significant loosening in coupling from 1999 to 2003, which may surprise educational policymakers. But the latter part of the NCLB era does depict an increase in micro-level coupling (from 2003), and it is possible that schools rebounded from the initial shock of NCLB mandates. Another reauthorization of ESEA means we are moving into a new policy era under the Every Student Succeeds Act (ESSA), and public schools may undergo another shift in coupling. Unlike IASA and NCLB, ESSA sends power back to the states to develop their own standards of accountability. Consequently, federal policy may become less important while school demographics and gendered leadership take center stage in molding schools’ micro-level coupling. Regardless, paying close attention to the trends of micro-coupling across public schools will be important for those interested in the organizational structure of schools. As principals gain more experience, schools become more loosely coupled. This finding implies that experience leads to looser coupling in schools. It is possible that more experienced principals recognize their teachers’ desire for loose coupling, and allow them more freedom. On average, female principals spend more time in the classroom prior to becoming principals, but female principals promote tighter couplings. Gender also intersects with federal policy eras in compelling ways. In recent waves, under NCLB, female principals were more likely than their male colleagues to manage tightly coupled schools. Furthermore, when female principals administrate in schools with a strong local government influence, the relationship to tight coupling is strengthened. These intersections emphasize how different levels of the school system work in concert with one another. The relationship between female principals and tighter coupling could signify that females are seeking legitimacy in their authority by enacting policies within the walls of their schools, especially if female principals are also seeking legitimacy from district officials when the district is tightly coupled to the school (meso-level coupling). Indeed, female principals have traditionally managed their schools using more participative strategies (Shakeshaft 1987). Tight meso-level coupling could prompt the district to hire a female principal who will actively impose policies within the school, although qualitative studies highlight how tighter couplings may undermine the legitimacy of the principal (with teachers) (Hallett 2007b, 2010). These are two possibilities that should be explored in future research. My analyses suggest a persistent role of race, but not class, in school coupling. The overarching goals of federal policies underscore the need to close the achievement gap for poor students and racial minorities, which could help explain the racial differences found here, but perhaps generate more questions regarding schools that serve poor students. These findings warrant more exploration, and future research should consider the processes under which principals and teachers shift their organizational structure as racial and socio-economic demographics shift. Longitudinal data with the same schools over time would be especially useful for scrutinizing this relationship. Perhaps unsurprisingly, urban schools have the tightest school-level coupling. Urban schools are often targeted in federal- and state-level policies. Recent qualitative studies on tighter couplings were conducted in urban schools and/or schools with higher proportions of minority students, and the findings presented here suggest that demographic characteristics are related to school-level coupling on a national level. These results are intriguing and merit further investigation, especially if federal- and/or local-level policies disproportionately affect some schools more than others. Finally, this study considers very specific dimensions of principal-teacher relationships, and coupling may tighten on some dimensions while loosening on others. Future research should consider other aspects of the principal-teacher relationship in analyzing how couplings exist and change within schools. While this work focuses on how couplings develop within schools, processes of coupling are embedded in other types of organizations, and this research expands our theoretical understanding of coupling. The public education system provides one type of organization to empirically test theories of coupling and neo-institutionalism, but schools are not the only organizations that experience these organizational processes. Similarly, the role of gender in positions of leadership is a critical component for constructing couplings. Understanding broader theoretical issues through the public education system will help scholars advance perspectives of coupling. Notes 1 Other characteristics of principals matter for management style—race could be an important factor, but teaching is an interesting organizational setting when it comes to gender and management. The overwhelming white demographic of teachers (86 percent) and principals (87 percent) makes understanding racial differences difficult in quantitative analysis. 2 The SASS data is a restricted-use dataset; this number is rounded in order to protect confidentiality. The number 34,940 is not the precise number of observations in this analysis, but it is the number reported in tables. 3 This indicates the percentage of teachers in these data, not the total percent of teachers in the school. Appendix Table A1. OLS Regression of Micro Coupling on Federal, State, Local, and Principal Characteristics   Model 1  Model 2  Model 3  Model 4  Socio-political factors          Year 1990  −0.052***  −0.052***  −0.057***  −0.049***  Year 1993  −0.075***  −0.074***  −0.082***  −0.061***  Year 1999  0.035  0.036  0.024  0.035  Year 2003  −0.076**  −0.075**  −0.091***  −0.117***  Year 2007  −0.031  −0.030  −0.046*  −0.075***  Year 2011  0.043*  0.043*  0.027  0.017  (Reference is 1987 (ANaR))          State characteristics          Index of high-stakes testing  0.011  0.011  0.011  0.011  Republican state (reference is Democrat)  −0.022  −0.022  −0.021  −0.021  Charter law  −0.002  −0.002  −0.003  −0.004  Local characteristics          Bonus    −0.011  −0.011  −0.010  Principal’s characteristics          Female principal (reference is Male)      0.061***  0.053**  Non-White principal (reference is White)      0.000  0.002  Highest degree—Doctorate      0.002  0.002  (reference is All other degrees)          Principal tenure (in years)      −0.003***  −0.003***  Teaching experience (in years) prior to principal      0.000  0.000  Interaction terms          Female principal x Year 1990        −0.031  Female principal x Year 1993        −0.078***  Female principal x Year 1999        −0.028  Female principal x Year 2003        0.073**  Female principal x Year 2007        0.073**  Female principal x Year 2011        0.029  Controls          Location          Northeast  −0.106*  −0.107*  −0.105*  −0.105*  Midwest  −0.166***  −0.166***  −0.163***  −0.162***  West  −0.105*  −0.106*  −0.109*  −0.107*  (reference is South)          Setting          Urban  0.180***  0.180***  0.173***  0.174***  Suburban  0.125***  0.125***  0.120***  0.121***  (reference is Rural)          School demographics          Free lunch percentage  0.000  0.000  0.000  0.000  Black percentage  0.003***  0.003***  0.003***  0.003***  Hispanic percentage  0.002***  0.002***  0.001***  0.001***  Asian percentage  −0.002**  −0.002**  −0.002**  −0.002**  American Indian percentage  0.001  0.001  0.001  0.001  (reference is White percentage)          Enrollment  0.000***  0.000***  0.000***  0.000***  Elementary school (reference is Secondary/Combined schools)  0.214***  0.214***  0.203***  0.202***  Percent of teachers in tested subject  0.002***  0.002***  0.002***  0.002***  Duplicate school  −0.007  −0.007  −0.005  −0.004  (Reference is Non-repeated school)          Constant  0.623***  0.623***  0.637***  0.635***  N = 40,360          R-squared  0.224  0.224  0.228  0.231    Model 1  Model 2  Model 3  Model 4  Socio-political factors          Year 1990  −0.052***  −0.052***  −0.057***  −0.049***  Year 1993  −0.075***  −0.074***  −0.082***  −0.061***  Year 1999  0.035  0.036  0.024  0.035  Year 2003  −0.076**  −0.075**  −0.091***  −0.117***  Year 2007  −0.031  −0.030  −0.046*  −0.075***  Year 2011  0.043*  0.043*  0.027  0.017  (Reference is 1987 (ANaR))          State characteristics          Index of high-stakes testing  0.011  0.011  0.011  0.011  Republican state (reference is Democrat)  −0.022  −0.022  −0.021  −0.021  Charter law  −0.002  −0.002  −0.003  −0.004  Local characteristics          Bonus    −0.011  −0.011  −0.010  Principal’s characteristics          Female principal (reference is Male)      0.061***  0.053**  Non-White principal (reference is White)      0.000  0.002  Highest degree—Doctorate      0.002  0.002  (reference is All other degrees)          Principal tenure (in years)      −0.003***  −0.003***  Teaching experience (in years) prior to principal      0.000  0.000  Interaction terms          Female principal x Year 1990        −0.031  Female principal x Year 1993        −0.078***  Female principal x Year 1999        −0.028  Female principal x Year 2003        0.073**  Female principal x Year 2007        0.073**  Female principal x Year 2011        0.029  Controls          Location          Northeast  −0.106*  −0.107*  −0.105*  −0.105*  Midwest  −0.166***  −0.166***  −0.163***  −0.162***  West  −0.105*  −0.106*  −0.109*  −0.107*  (reference is South)          Setting          Urban  0.180***  0.180***  0.173***  0.174***  Suburban  0.125***  0.125***  0.120***  0.121***  (reference is Rural)          School demographics          Free lunch percentage  0.000  0.000  0.000  0.000  Black percentage  0.003***  0.003***  0.003***  0.003***  Hispanic percentage  0.002***  0.002***  0.001***  0.001***  Asian percentage  −0.002**  −0.002**  −0.002**  −0.002**  American Indian percentage  0.001  0.001  0.001  0.001  (reference is White percentage)          Enrollment  0.000***  0.000***  0.000***  0.000***  Elementary school (reference is Secondary/Combined schools)  0.214***  0.214***  0.203***  0.202***  Percent of teachers in tested subject  0.002***  0.002***  0.002***  0.002***  Duplicate school  −0.007  −0.007  −0.005  −0.004  (Reference is Non-repeated school)          Constant  0.623***  0.623***  0.637***  0.635***  N = 40,360          R-squared  0.224  0.224  0.228  0.231  Note: Schools are clustered in states. Data are from the National Center for Education Statistics (NCES) Schools and Staffing Survey (SASS) and, following NCES convention, I have rounded sample size numbers to the nearest 10 in order to protect the identities of respondents. Given the large number of cases, models were examined using a random 10 percent subsample, and the results are substantively similar. The sample uses all traditional schools in the 1987, 1990, 1993, 1999, 2003, 2007, and 2011 waves of SASS data that have valid information on my independent, dependent, and control variables. I exclude private, charter, alternative, and non-traditional schools. *p < 0.05 **p < 0.01 ***p < 0.001 About the Author Maria Paino is an Assistant Professor in the Department of Sociology, Anthropology, Social Work, and Criminal Justice at Oakland University. Her research focuses on inequalities and organizational processes particularly in educational and clinical contexts. References Acker, Joan. 1990. “ Hierarchies, Jobs, Bodies: A Theory of Gendered Organizations.” Gender & Society  4( 2): 139– 58. Google Scholar CrossRef Search ADS   Aldrich, Howard E., and Martin Reuf. 2006. Organizations Evolving . London: Sage. Bidwell, Charles E. 2001. “ Analyzing Schools as Organizations: Long Term Permanence and Short Term Change.” Sociology of Education  ( Extra Issue): 100– 114. Bossert, Steven T., David C. Dwyer, Brian Rowan, and Ginny V. Lee. 1982. “ The Instructional Management Role of the Principal.” Educational Administration Quarterly  18( 3): 34– 64. Google Scholar CrossRef Search ADS   Carnoy, Martin, and Susanna Loeb. 2002. “ Does External Accountability Affect Student Outcomes? A Cross-State Analysis.” Educational Evaluation and Policy Analysis  24( 4): 305– 31. Google Scholar CrossRef Search ADS   Charters, W. W. Jr., and T. D. Jovick. 1981. “The Gender of Principals and Principal/Teacher Relations in Elementary Schools.” In Educational Policy and Management: Sex Differentials , edited by P. A. Schmuck, W. W. Charters Jr., and R. O. Carlson, 307– 331. New York: Academic Press. Coburn, Cynthia E. 2004. “Beyond Decoupling: Rethinking the Relationship between the Institutional Environment and the Classroom.” Sociology of Education  77( 3): 211– 44. Google Scholar CrossRef Search ADS   Cochran-Smith, Marilyn, and Susan L. Lytle. 2006. “ Troubling Images of Teaching in No Child Left Behind.” Harvard Educational Review  76( 4): 668– 97. Google Scholar CrossRef Search ADS   Cohen, Michael D., and James G. March. 1974. Leadership and Ambiguity . New York: McGraw-Hill. Darling-Hammond, Linda. 2007a. “ Race, Inequality and Educational Accountability: The Irony of ‘No Child Left Behind.’” Race Ethnicity and Education  10( 3): 245– 60. Google Scholar CrossRef Search ADS   Darling-Hammond, Linda. 2007b. “Evaluating ‘No Child Left Behind.’” The Nation 21(May):1–7. Davies, Scott, Linda Quirke, and Janice Aurini. 2006. “The New Institutionalism Goes to the Market: The Challenge of Rapid Growth in Private K-12 Education.” In The New Institutionalism in Education , edited by H. D. Meyer and B. Rowan, 103– 22. Albany, NY: SUNY Press. Desimone, Laura. 2013. “ Teacher and Administrator Responses to Standards-Based Reform.” Teachers College Record  115( 8): 1– 53. Diamond, John B. 2007. “ Where the Rubber Meets the Road: Rethinking the Connection between High-Stakes Testing Policy and Classroom Instruction.” Sociology of Education  80( 4): 285– 313. Google Scholar CrossRef Search ADS   Diamond, John B. 2012. “ Accountability Policy, School Organization, and Classroom Practice: Partial Recoupling and Educational Opportunity.” Education and Urban Society  44( 2): 151– 82. Google Scholar CrossRef Search ADS   Eagly, Alice H., and Blair T. Johnson. 1990. “ Gender and Leadership Style: A Meta-Analysis.” Psychological Bulletin  108: 233– 56. Google Scholar CrossRef Search ADS   Eagly, Alice H., and Steven J. Karau. 2002. “ Role Congruity Theory of Prejudice Towards Female Leaders.” Psychological Review  109( 3): 573– 98. Google Scholar CrossRef Search ADS PubMed  Elmore, Richard F., Charles Abelman, and Susan H. Furman. 1996. “The New Accountability in State Education Reform: From Process to Performance.” In Holding Schools Accountable , edited by H. F. Ladd, 65– 98. Washington, DC: Brookings Institution. Fauth, Gloria C. 1984. “ Women in Educational Administration: A Research Profile.” Educational Forum  49( 1): 65– 79. Google Scholar CrossRef Search ADS   Firestone, W. A. 1985. “The Study of Loose Coupling: Problems, Progress, and Prospects.” In Research in Sociology of Education and Socialization , vol. 5, edited by A. Kerckhoff, 3– 30. Greenwich, CT: JAI Press. Gamoran, Adam, and Robert Dreeben. 1986. “ Coupling and Control in Educational Organizations.” Administrative Science Quarterly  31: 612– 32. Google Scholar CrossRef Search ADS   Grissom, Jason A., Demetra Kalogrides, and Susanna Loeb. 2013. “Strategic Staffing: Examining the Class Assignments of Teachers and Students in Tested and Untested Grades and Subjects.” Paper presented at the Association for Education Finance and Policy, March 14, New Orleans, LA. Hallett, Tim. 2007a. “ Between Deference and Distinction: Interaction Ritual through Symbolic Power in an Educational Institution.” Social Psychology Quarterly  70( 2): 148– 71. Google Scholar CrossRef Search ADS   Hallett, Tim. 2007b. “The Leadership Struggle: The Case of Costen Elementary School.” In Distributed Leadership in Practice , edited by J. Spillane and J. B. Diamond, 85– 105. New York: Teachers College Press. Hallett, Tim. 2010. “ The Myth Incarnate: Recoupling Processes, Turmoil, and Inhabited Institutions in an Urban Elementary School.” American Sociological Review  75( 1), 52– 74. Google Scholar CrossRef Search ADS   Hanson, Mark. 1977. “ Beyond the Bureaucratic Model: A Study of Power and Autonomy in Educational Decision-Making.” Interchange  7( 2): 27– 38. Google Scholar CrossRef Search ADS   Ingersoll, Richard M. 1996. “ Teachers’ Decision-Making Power and School Conflict.” Sociology of Education  69( 2): 159– 76. Google Scholar CrossRef Search ADS   Ingersoll, Richard M. 2003. Who Controls Teachers’ Work? Power and Accountability in American Schools . Cambridge, MA: Harvard University Press. Jennings, Jennifer L. 2010. “ School Choice or Schools’ Choice? Managing in an Era of Accountability.” Sociology of Education  83( 3): 227– 47. Google Scholar CrossRef Search ADS   Johnson, Shirley, Steve Busch, and John R. Slate. 2008. “Leadership Behaviors of School Administrators: Do Men and Women Differ?” International Journal of Educational Leadership Preparation 3(3): 1– 11. Le Floch, Kerstin Carlson, James E. Taylor, and Kerri Thomsen. 2006. “ Implications of Nclb Accountability for Comprehensive School Reform.” Journal of Education for Students Placed at Risk  11( 3– 4): 353– 66. Google Scholar CrossRef Search ADS   Lee, Valerie E., Robert F. Dedrick, and Julia B. Smith. 1991. “ The Effect of the Social Organization of Schools on Teachers’ Efficacy and Satisfaction.” Sociology of Education  64( 3): 190– 208. Google Scholar CrossRef Search ADS   Lee, Valerie E., Julia B. Smith, and Madalyn Cioci. 1993. “ Teachers and Principals: Gender-Related Perceptions of Leadership and Power in Secondary Schools.” Educational Evaluation and Policy Analysis  15( 2): 153– 80. Google Scholar CrossRef Search ADS   Lortie, Dan C. 2002. Schoolteacher: A Sociological Study . Chicago: University of Chicago Press. March, James G., and Johan P. Olsen. 1976. Ambiguity and Choice in Organizations . Bergen: Universitetsforlaget. Mathis, William J. 2003. “ No Child Left Behind: Costs and Benefits.” The Phi Delta Kappan  84( 9): 679– 86. Google Scholar CrossRef Search ADS   McDermott, Kathryn A. 2011. High Stakes Reform: The Politics of Educational Accountability . Washington DC: Georgetown University Press. Meyer, John W., and Brian Rowan. 1977. “ Institutionalized Organizations: Formal Structure as Myth and Ceremony.” American Journal of Sociology  83: 340– 63. Google Scholar CrossRef Search ADS   Meyer, John W., and Brian Rowan. 1978. “The Structure of Educational Organizations.” In Environments and Organizations , edited by M. W. Meyer, 78– 109. San Francisco: Jossey-Bass. National Center for Education Statistics. 1994. “Public and Private School Principals: Are There Too Few Women?” Washington, DC: US Department of Education. National Center for Education Statistics. 2013. “Characteristics of Public and Private Elementary and Secondary School Principals in the United States.” Washington, DC: US Department of Education. Office of the Federal Register. 2013. “U.S. Electoral College.” Washington, DC: US National Archives and Records Administration. http://www.archives.gov/federal-register/electoral-college/about.html. Orlich, Donald C. 2004. “ No Child Left Behind: An Illogical Accountability Model.” The Clearing House  78( 1): 6– 11. Google Scholar CrossRef Search ADS   Orton, J. Douglas, and Karl E. Weick. 1990. “ Loosely Coupled Systems: A Reconceptualization.” Academy of Management Review  15( 2): 203– 23. Pfeffer, Jeffrey. 1981. Power in Organizations . Boston: Pitman. Pierce, Jennifer. 1995. Gender Trials . Berkeley: University of California Press. Pitner, Nancy J. 1981. “Hormones and Harems: Are the Activities of Superintending Different for a Woman?” In Educational Policy and Management: Sex Differentials , edited by P. A. Schmuck, W. W. Charters Jr., and R. O. Carlson, 273– 295. New York: Academic Press. Price, Heather. 2012. “ Principal-Teacher Interactions: How Affective Relationships Shape Principal and Teacher Attitudes.” Educational Administration Quarterly  48( 1): 39– 85. Google Scholar CrossRef Search ADS   Renzulli, Linda A., and Vincent J. Roscigno. 2005. “ Charter School Policy, Implementation, and Diffusion across the United States.” Sociology of Education  78( 4): 344– 65. Google Scholar CrossRef Search ADS   Renzulli, Linda A., Heather Macpherson Parrott, and Irenee R. Beattie. 2011. “ Racial Mismatch and School Type: Teacher Satisfaction and Retention in Charter and Traditional Public Schools.” Sociology of Education  84( 1): 23– 48. Google Scholar CrossRef Search ADS   Rowan, Brian. 1981. “The Effects of Institutionalized Rules on Administrators.” In Organizational Behavior in Schools and School Districts , edited by S. B. Bacharach, 47– 75. New York: Praeger. Rowan, Brian. 1990. “ Commitment and Control: Alternative Strategies for the Organizational Design of Schools.” Review of Research in Education  16: 359– 89. Sauder, Michael, and Wendy Nelson Espeland. 2009. “ The Discipline of Rankings: Tight Coupling and Organizational Change.” American Sociological Review  74: 63– 82. Google Scholar CrossRef Search ADS   Seyfarth, John T., and William A. Bost. 1982. “ Prejudice in the Principal’s Office: Administrator Attitudes toward Older Teachers.” High School Journal  65( 5): 157– 62. Shakeshaft, Charol. 1987. Women in Educational Administration . Newbury Park, CA: Sage. Spillane, James P., and Patricia Burch. 2006. “The Institutional Environment and Instructional Practice: Changing Patterns of Guidance and Control in Public Education.” In The New Institutionalism in Education , edited by H.-D. Meyer and B. Rowan, 87– 102. Albany: State University of New York Press. Spillane, James P., John B. Diamond, Patricia Burch, Tim Hallett, Loyiso Jita, and Jennifer Zoltners. 2002. “ Managing in the Middle: School Leaders and the Enactment of Accountability Policy.” Educational Policy  16( 5): 731– 62. Google Scholar CrossRef Search ADS   Spillane, James P., Tim Hallett, and John B. Diamond. 2003. “ Forms of Capital and the Construction of Leadership: Instructional Leadership in Urban Elementary Schools.” Sociology of Education  76( 1): 1– 17. Google Scholar CrossRef Search ADS   Spillane, James P., Leigh Mesler Parise, and Jennifer Zoltners Sherer. 2011. “ Organizational Routines as Coupling Mechanisms: Policy, School Administration, and the Technical Core.” American Educational Research Journal  48( 3): 586– 619. Google Scholar CrossRef Search ADS   Tyler, Tom R. 2006. “ Psychological Perspectives on Legitimacy and Legitimation.” Annual Review of Psychology  57: 375– 400. Google Scholar CrossRef Search ADS PubMed  US Department of Education. 2013. “States Granted Waivers from No Child Left Behind Allowed to Reapply for Renewal for 2014 and 2015 School Years.” Washington, DC. Valli, Linda, and Daria Buese. 2007. “ The Changing Roles of Teachers in an Era of High-Stakes Accountability.” American Educational Research Journal  44( 3): 519– 58. Google Scholar CrossRef Search ADS   Weeden, L. Darnell. 2005. “Essay: Does the No Child Left Behind Law (NCLBA) Burden the States as an Unfunded Mandate under Federal Law [article]." Thurgood Marshall Law Review 31( 2): 239– 252. Weick, Karl. 1976. “ Educational Organizations as Loosely Coupled Systems.” Administrative Science Quarterly  21: 1– 19. Google Scholar CrossRef Search ADS   Weiss, Carol H., and Joseph Cambone. 1994. “ Principals, Shared Decision Making, and School Reform.” Educational Evaluation and Policy Analysis  16( 3): 287– 301. doi:10.2307/1164401. Google Scholar CrossRef Search ADS   Williams, Christine L. 1993. Doing “Women’s Work”: Men in Nontraditional Occupations . Newbury Park: Sage. Google Scholar CrossRef Search ADS   Wingfield, Adia Harvey. 2009. “ Racializing the Glass Escalator: Reconsidering Men’s Experiences with Women’s Work.” Gender and Society  23( 1): 5– 26. doi:10.2307/20676747. Google Scholar CrossRef Search ADS   Young, Viki M. 2006. “ Teachers’ Use of Data: Loose Coupling, Agenda Setting, and Team Norms.” American Journal of Education  112( August): 521– 48. Google Scholar CrossRef Search ADS   Zelditch, Morris Jr. 2001. “Theories of Legitimacy.” In The Psychology of Legitimacy: Emerging Perspectives on Ideology, Justice, and Intergroup Relations , edited by J. Jost and B. Major, 33– 53. Cambridge: Cambridge University Press. Author notes I would like to thank Linda Renzulli, Ashley Barr, Elizabeth DeBray, Jeremy Reynolds, Rebecca Boylan, and Dennis Condron for their comments on earlier drafts. Please feel free to contact me at paino@oakland.edu. © The Author 2017. Published by Oxford University Press on behalf of the University of North Carolina at Chapel Hill. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.

Journal

Social ForcesOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off