1
The Teachers’ Union as a Knowledge Network:
Evidence from United States Public Schools
John E. McCarthy
Rutgers University & University of Pennsylvania
Saul A. Rubinstein
Rutgers University
June 2014
Under Review at Industrial and Labor Relations Review
Funding for this research was provided by the Bill and Melinda Gates Foundation.
Abstract
School districts house considerable intellectual resources that often go underutilized because schools are insulated from one another – operating, for the most part, as islands. Education scholars have argued that teachers and students stand to benefit from boundary spanning networks throughout the school district because these ties have the potential to introduce external perspectives and break down conformity pressures. The present study examines labor management partnerships in general, and school union representatives in particular, as potential sources of knowledge diffusion between schools. Webuild from existing theory, as well as multiple years of qualitative and quantitative data collected in a high partnership school district,to develop two hypotheses concerning union representatives’ external social capital and the receipt and application of external knowledge by teachers in schools. We test these hypotheses by combining two data sources, obtained from union representatives and, separately, the teachers whose interests they represent. Our results suggest that school union representatives can bridge important knowledge boundaries between schools but that a school’s capacity to utilize external knowledge depends on the level of labor-management partnership in place.
Introduction
One of the most pressing policy issues over the past three decades is how to encourageeffective, lastingimprovement intoclassrooms (Evans, 2001). Since the publication of A Nation at Risk in the early 1980s, schools in the United Stateshave been subject to intensifying scrutiny by policymakers, legislators and the general public. The schoolteacher has found herself on the defensive, subject to educational fads and control mechanisms that are determined, and overseen, by bureaucrats far removed from her classrooms and students. Race to the Top and No Child Left Behind with their focus on high stakes standardized testing, value-added teacher evaluations, expansion of market-based solutions like Charter Schools, and the more recent shift toward common core standards provide examples. While some teachers have embraced these changes, the majority has not (Evans2001; Kim and Orfield 2004; SundermanTracey, 2004). Scholars are also cautious of top-down reform efforts – urging that the recipe for lasting school improvement stems from empowerment and meaningful collaboration among teachers (Rubinstein and McCarthy, 2012; Rubinstein and McCarthy, 2014).
In addition to collaboration within schools (Leana and Pil, 2006; Louis and Marks 1998; Rubinstein and McCarthy 2014), numerous scholars have encouragedgreater networking between schools in school districts as a way to promote school improvement (seeEaker, DuFour and Eaker 2008; Fullan 2010; Stoll, Bolam, McMahon, Wallance and Thomas 2006). According to this latter perspective, school districts house considerable intellectual resources throughout their schools but these resources oftengo under-utilizedbecause schools, as traditionally structured, are highly insulated from one another (Hargreaves and Giles 2003). Successful strategies or innovations being employed by one school are often stuck – inaccessible by the other schools in the district. Thus, increasing schools’ external connectedness may broaden perspectives and weaken the complacency and conformity pressures found in highly insulated groups(see Stoll and Seashore-Lewis 2007: 7).
As of yet, there has beenlittle research into the effects of inter-school collaboration. Nor have researchers examined viableworkplace institutions that support it. Ourstudycombines qualitative and quantitative evidence to examine antecedents and consequences of inter-school collaboration. Specifically, our research in a school district with a union-management partnership arrangement sustained over 15 years highlights the teacher union as an effective instrument for school changethroughschool union representatives’ knowledge sharing connections throughout the school district. We find not only that school union representatives are heavily engaged in knowledge sharing communications between schools but also that these connectionshave direct implications for teachers – in particular, teachers’ awareness and application of external knowledge. However, wealso find some evidence that external knowledge goes further in some schools relative to others: Our data suggest that schools may benefit more from union representatives’external ties when strong school-level labor-management partnerships – characterized by shared decision-making and collaboration– are in place.
This study makes several important contributions, both practical and theoretical. While there are recognized benefits to boundary spanning in organizations (e.g., Ancona1990), less attention has been given to workplace institutions that support boundary-spanning activities (e.g., Mehra, Kilduff and Brass 2001) – an oversight that generalizes to public schools. Indeed, despite the potential benefits of inter-school collaboration, some scholars are hesitant about increasing collaborative obligations placed on teachers because their jobs are demanding and their time limited (Little 2005). In order for inter-school knowledge sharing to successfullytake hold,therefore, schools may need designated boundary-spanners who take the responsibility for building and maintaining external connections and redistributing resources back into their school. Our research suggests that school union representatives can fill this void organically by expandingthe scope of existing support services.
This study also makes contributions to industrial relations research and theory. Scholars have lamented unions’ historically weak contributions to social capital (Cornwell and Harrison 2004; Jarley 2005) – i.e., interpersonal relationships that confer or potentiate value to members (Adler and Kwon 2002). Our research suggests thatunions can enhance organizational outcomes through their contributions to “boundary-spanning” tiesthat are vital for disseminating knowledge and resources in large organizations (Oh, Labianca and Chung 2004).Organizations often encounter challenges as they attempt to move high-level partnerships to lower levels (Cutcher-Gershenfeld, Kochan, and McKersie 1988; Heckscher and Schurman 1997; Rubinsteinand Kochan 2001).In addition to showing some potential benefits of union-created social capital, our results suggest that union representatives’ external ties may bear more influence – i.e., get more millage – in high partnership settings relative to low partnership settings. We believe that this highlights a challenge of optimizing on labor-management partnerships in large, decentralized organizations.
Theoretical Background
Social Capital
Social capital theory is based on idea that interpersonal relationships serve as powerful conduits for information and resource transfer and help to mobilize social benefits. The concept has gained a strong footing in the organizational sciences, given that relationships powerfully affect opportunity structures within organizations (Burt 1992; Seibert, Kraimer, and Liden 2001) as well coordination capabilities and the sharing and recombination of knowledge resources between individuals and groups (Burt 2004; Tsai and Goshal 1998). Drawing from the sociological theories laid down by Coleman (1990) and Burt (1992), this literature has highlighted the importance of cohesive, trusting ties among group members, which facilitate local knowledge sharing and recombination (Collins and Smith 2006) and help to mobilize efforts around a common direction or innovative pursuit (Obstfelt 2005) as well as more distal external linkages that connect individuals and groups to a broader pool of resources located throughout their organization system (Ancona 1990).
Social Capital and Labor Unions
Industrial relations scholarship has recognized the importance of social capital, though primarily from the standpoint of building mobilization capacity and revitalizing organized labor (Heckscher 1998; Jarley 1995). For example, Heckscher (1988) was hopeful that identity groups would coalesce with one another around more general issues relating to workplace rights. He was optimistic that disparate identities groups would find natural allies with one another and would unite in supporting one another in collective action. These identity groups have had successes[1] but thus far have not matched the strength or durability of traditional labor unions. Furthermore, labor unions have approached these new forms of social organization tepidly, at best(Heckscher and McCarthy2015). Craft unions have been particularly resistant to outsiders. The Service Employees’ union (SEIU) has perhaps gone the furthest in aligning itself with beleaguered identity groups but in general union members and union leadership have remained significantly insulated – preoccupied, for the most part, with local working conditions and insurance provisions (e.g., Cornwell and Harrison 2004).
Some scholarshave also noted the decline of social capital within labor organizations (Jarley 2005). Early modes of organizing began as tightly bonded mutual-aid networks in which working-class craftsmen pulled together to support labor standards while offering assistance to the sick and unemployed and their families (Cobble 1990; Beito 1999; Bacharach et al. 2001; Jarley 2005). These groups were heavily socially oriented, with loyalties fortified by dense,stable bonds that conferred privileged resource accessto in-group members (Heckscher and McCarthy2015). The mutual aid model of organized labor was significantly disrupted in the early twentieth century by the rise of large industries and economies of scale and as business and government subsumed social welfare functions (Jarley 2005). Bacharach and colleagues (2001) note that collective bargaining and political lobbying, carried out by a small group of representative staff, took the place of social community and shared obligations to provide mutual aid. Compelling evidence of weakening solidarity is found in declining instances of mass action observed throughout the OECD area over the last thirty years (see Heckscher and McCarthy2015).
While pessimism over the labor movement’s building ofsocial capital may be well founded, most scholarship hasfocused on labor’s capacity to amass leverage and political influence over recalcitrant employers. Scholars have given less attention to the value that unions can add to organizational effectiveness through their contributions to social capital (Rubinstein 2000; Rubinstein 2001). The literature on value-added unionism notes that strong labor-management partnerships can enhance organizational outcomes via institutions and union-based networks that facilitate communication and broaden collaboration and participation in decision-making among employees. Employees are more likely to develop relationships with one another if they have opportunities to interact throughout the workday (Whitener, Brodt, Korsgaard and Werner 1998) and team-based work designs thus significantly broaden social exposure. Conceivably, these high involvement work systems, facilitated through partnership, should increase social capital to the extent that they bring workers into dialogue around substantive issues and create a sufficient level of employment stability to enable shared histories and trust.
Rubinstein (2001: 589) has argued that labor-management partnerships can contribute to organizational social capital and in particular a “kind of communication and coordination network needed to produce the flexibility and responsiveness required of today’s ‘high-performance’ work systems.” Consistent with this argument, Rubinstein (2000) found that the network infrastructureincluding communications, union-management alignment, and balance of time spent managing people and production explained 30 percent of the variation in first-time quality and 53 percent of the variation in quality improvement. However, the most significant contribution to quality and quality improvement came from communications which were stronger for union represented managers relative to non-represented managers, suggesting that labor unions can contribute significantly to firm effectiveness in ways unattainable in non-union firms because of the network infrastructure the union can produce.
More recently, Rubinstein and McCarthy (2012) showed that labor management partnerships in school districts reinforce and are reinforced by continuous communication and coordination between management and elected union leaders. Analyzing communication patterns between educators within schools, Rubinstein and McCarthy (2014) found evidence that school-level partnerships can foster higher levels of collaboration among teaching faculty, which can result in higher student performance outcomes.
The Union as a Boundary Spanning Network
With some exceptions (Rubinstein2000; 2001; Rubinstein and McCarthy 2012. Rubinstein and McCarthy2014), labor-management partnerships’ contributions to organizational effectiveness throughsocial capital are seldom examined and poorly understood. We address this gap by examiningalocal union’s role in fostering social capital between schools within a school district. We conducted over 40 interviews ina high partnership school district from 2008 to 2013. We interviewed principals, teachers and union representatives across schools. On numerous occasions, we interviewed superintendents, school board members, central office administrators, human resource leaders, and unionpresidents and executive board members (there was leadership succession over the period of our study). We sat in on meetings and district-sponsored events, including sessions, outlined below, that hosted joint presentations by school principals and school union representatives. We conducted social network analysis to capture relationships not only within but also between the district’s schools. Collectively, our evidence suggested thatunion representatives in a higher partnership school district were playing a critical role in disseminating knowledge between schools.The following incorporatesexisting theory and research, as well as our own qualitative evidence, to inform two hypotheses relating to union representatives’ network and their implications for teachers in schools – specifically, we focus on the extent to which teachers obtainknowledge that originated in other schools in the district and whether they apply this knowledge to improve their instruction.
Hypotheses
Boundary Spanning
Organizational boundary spanning creates social capital for organizations by connecting groups with different types of knowledge and resources (Burt, 2005). These connections can facilitate coordination and innovation (Tushman 1977), particularly in knowledge-intensive settings that are functionally interdependent and benefit from the transfer and recombination of diverse ideas and perspectives (Aldrich and Herker 1977; Tushman 1977; Roberts and O’Reilly 1979; Allen 1984; Gladstein 1984; Ancona and Caldwell 1988; Ancona 1999). Its benefits are complimentary to the theory of “structural holes”, introduced and elaborated by Burt (1992), which views non-redundant external linkages as a source for timely information and resource access and intellectual diversity. Although structural holes and boundary spanning have strong conceptual ties, and co-vary empirically (Fleming and Waguespack 2007), the two have traditionally been different in their focus. The focus of structural holes is usually competitive advantage for the individual who spans across social boundaries. By contrast, the boundary-spanning literature has focused on externally networked intermediaries that redirect external information and resources for a collective benefit(Fleming and Waguespack 2007).
Boundary spanning’s benefitshave been supported by numerous studies and at multiple levels of analysis. McEvily and Zaheer (1999) found that Michigan manufacturers who held a greater number of non-redundant sources of advice outside of the firm had better access to competitive ideas. Geletkanycz and Hambrick (1997) found that companies perform better when top-managers held boundary-spanning relationships outside of their firm and industry. Baum, Calabrese and Silverman (2000) found that Canadian biotechnology companies with more heterogeneous networks experienced faster revenue growth and were better able to obtain patents than organization with homogeneous, overlapping networks. Within organizations, Rodan and Galunic (2004) found that that heterogeneous sparse knowledge predicted innovation. Ancona's (1990) research supported a link between externally oriented boundary-spanning activities (informing, parading and probing) and team performance. Baldwin, Bedell and Johnson (1997) found that MBA teams whose members were better networked performed better on a class project. Hansen, Podolny and Pfeffer (2001) found that teams completed task assignments more quickly when they held more non-redundant contacts outside of the team. Hansen (1999) found that weak, diverse networks between team members were instrumental in helping teams search for new knowledge.
As cited, several education researchers have suggested that school districts may benefit from knowledge sharing between their schools (see Hargreaves and Giles 2003; Stoll, Bolam, McMahon, Wallance and Thomas 2006; Eaker, DuFour and Eaker 2008; Fullan 2010). Inter-school boundary spanning connections are seen as important because they have the potential to broaden the boundaries of knowledge exchange and combination and increase resource access and diffusion between schools. As in any organization or work group, schools can become fixed on a narrow set of perspectives, which often limits the school’s potential for meaningful change. Exposure to the practices and learning techniques employed at other sites can foster meaningful innovation1 and professional improvement (e.g., Fullan 2010). It is for these reasons that Stoll and Seashore-Lewis (2007: 7) write that “without due attention to fostering ties outside the school, strong professional communities can, paradoxically, become a barrier to change.” Hargreaves and Giles (2003: 134) emphasize similarly that school districts must successfully bring “together the knowledge, skills and dispositions of teachers in a school or across schools to promote shared learning and improvement.”
Although there is a general understanding that organizations and organizational institutions affect the formation and dissolution of ties (Nahapiet and Ghoshal 1998), the antecedents to organizational boundary spanning has received comparatively little attention from organizational researchers(Mehra, Kilduff and Brass 2001). Nahapiet and Ghoshal (1998) note that organizations contribute to the formation of social capital by bringing people together for extended periods of time, by making individuals interdependent with one another, by encouraging them to interact in reciprocal exchanges, and by providing an element of closure (by way of organizational and sub-group boundaries), which facilitates norms, social identities and trust. In a related vein, scholars recognize that organizational initiatives and HR practices – including, for examples, cross-functional teams, job rotation programs, and individual and group pay policies –powerfully direct the nature of workers’ social connections as well as their efficacy from the standpoint of advancing organizational objectives (Baron and Pfeffer 1994; Collins and Smith 2006; Evans and Davis 2005).