Beyond the Labor Market Paradigm : A Social Network Perspective on Teacher Recruitment and Retention 1

This article identifies limits of the dominant labor market perspective (LMP) in research on teacher recruitment and retention and describes how research that incorporates a social network perspective (SNP) can contribute to the knowledge base and development of teacher education, staffing, and professional development approaches. A discussion of current literature on teachers' social networks and a case example of social network perspective research highlight how such research reveals complex social factors that shape teachers' workplace experiences and show the ways in which the labor market perspective tends have a "recruitment-heavy" focus. The article describes how the social network perspective allows researchers to study social workplace concerns from both macro and micro perspectives. This approach can broaden the current focus on recruitment to a more comprehensive understanding of recruitment and retention.


Introduction
Since the early 1990's a wide body of research and literature has addressed the problem of teacher attrition.Labor market concepts dominate the theoretical frameworks of research in this field, often omitting the social details of factors related to teachers' working conditions that could contribute to a fuller understanding of these issues (Borman & Dowling, 2008).A social network perspective (hereafter referred to as the SNP) is an emerging research paradigm in education that focuses on the exchange of information, resources, support, and trust through social networks.It offers new insights and conceptual frameworks for understanding the "revolving door" problem in teacher recruitment and retention (Ingersoll, 2001) and may help to explain and develop effective policy solutions.
Alongside our existing knowledge base, SNP research can help to provide a comprehensive picture of teacher recruitment and retention policies.The problem of teacher turnover is linked to issues of teacher recruitment, induction, and long-term retention efforts.Studies framed by the labor market perspective (referred to here as the LMP), the primary research paradigm in the field, tend to focus heavily on the recruitment end of the problem, as the majority of LMP research is on financial and human capital incentives to attract and retain high quality teachers.Complex social factors are widely missing in LMP research.However, these factors-often described in LMP as aspects of teachers' "working conditions"-influence teachers' commitment, professional learning, and ability to cope with change, and they are key to retaining teachers (Hanushek, Kain, & Rivkin, 2004).Further, scholars largely agree that the root of the "teacher shortage" problem lies in the retention of quality teachers (Ingersoll & Smith, 2003).The SNP addresses this missing link in the research on teacher recruitment and retention by focusing on relationships and the interchange of social capital.
In this article I examine the SNP and provide specific examples of how it can address some of the missing links in current research and policy.I begin with a description of the contributions and limits of research and policy framed by the LMP.Next, I review SNP research and literature, focusing on its current role in research on teachers.Following these literature reviews, through one case I demonstrate how one can use SNP to uncover results of teacher hiring policies that were not fully revealed from the LMP.Through this example I illustrate the ways that SNP can pose significantly new questions and perspectives for educational research and policy.

The Labor Market Perspective: A Front-End Focus
The LMP is based on an economic perspective on labor, framed with the market concepts of supply and demand.The principle of teacher supply is a tenet of this framework: teachers' decisions to apply to, accept jobs at, and continue working at a school are based on the perceived benefit of working at the school (financial, social, or otherwise) as compared with other opportunities (Guarino, Sanibanez, & Daley, 2006).LMP research examines teachers' reported satisfaction in areas such as salary, benefits, working conditions, and efficacy and makes associations between these findings and attrition data.This perspective has primarily focused on the individual teacher (Ingersoll, 2001).
The LMP framework is useful in evaluating the effects of three approaches that aim to attract high quality teachers: offering financial incentives, changing entry requirements, and developing human capital.The majority of teacher recruitment and retention policies incorporate some aspect of these approaches.This section examines how particular policies that used these three approaches were framed and studied by LMP, highlighting two trends.First, the policies have the tendency to focus on front-end attractiveness solutions rather than long-term retention strategies.Second, LMP has limitations in examining the social aspects of the policies and approaches.
Financial incentives include teachers' base salary, merit pay, bonuses, and tuition reimbursement.While there is some evidence that schools have greater success in recruiting teachers and improving student achievement through higher pay (Loeb & Page, 2000), the role of teacher salary and specifically bonus pay is less influential in retaining teachers than teachers' workplace experiences and access to support (Ingersoll, 2001).For example, Liu, Johnson, & Peske (2004) studied the effects of the Massachusetts $20,000 bonus program over several years, and cited only "distant effects" of the bonus on teachers' decisions to stay in teaching.They concluded that the policy "focused too narrowly on recruitment and not enough on retention" (p.234).They attributed most teachers' decisions to leave on the lack of structures and norms in the school to support teacher collaboration.
Another idea to attract and retain high quality teachers has been to change entry requirements for the teaching profession (Baker & Dickerson, 2006;Darling-Hammond & Sykes, 2003;Podgursky, Monroe, & Watson, 2004).The entry requirements for Teach for America (TFA), a national alternative teacher training program, require little to no previous teacher education or experience.In TFA, college graduates accepted into the program take a six-week summer course in advanced preparation for teaching and then continue coursework over one or two years as they teach.Boyd, Grossman, Lankford, Loeb, and Wykoff (2006) examined TFA teachers in New York City and found that by their fourth year, 81% of TFA teachers left their schools or the teaching profession altogether, compared with 44% of traditional-entry teachers.The authors note, "in many respects, alternative routes are designed as recruitment strategies" (p.212).The focus for TFA and similar programs is on recruiting academically strong candidates; to provide incentives for such applicants the appointment is conceptualized as a short-term (two to three year) position rather than a long-term career.Thus, this type of approach helps to recruit many new, enthusiastic teachers but offers fewer solutions for retention.
Within LMP research, the concept of human capital has been used to frame issues such as the attractiveness of schools and working conditions.Human capital is the level of formal professional training and education acquired by an individual (Becker, 1962).The LMP framework is challenged, however, in examining and discussing the social particularities of human capital development in teaching.One example of the use of the notion of human capital to devise hiring practices is the implementation of school reconstitution, in which a large percentage of a school faculty are replaced by teachers with higher credentials (Hess, 2003); this practice is the focus of the case example presented later in this article.
Another example of the LMP focus on human capital is research on the teacher induction movement.Teacher induction has earned a major role in addressing the problem of teacher turnover, due in large part to LMP research.However, as is the case in the policies described above, the emphasis on induction has developed as a front-end strategy focused on what Cochran Smith and Lytle (1999) call "knowledge for practice" (i.e.specific skills) rather than supporting teachers "knowledge of practice" (e.g., teacher inquiry and collaboration) through an ongoing, comprehensive approach to recruitment and retention.Many states now mandate an induction program for new and incoming teachers, yet the majority of these induction programs do not last beyond the first year (Furtwengler, 1995), barely enough time for new teachers to nurture and sustain the kinds of supportive networks that can keep them committed to the profession.
The strength in LMP research is in measuring the attractiveness of policies or programs to particular types of teacher; however, it does less to explain and examine the role of relationships, social capital, and social networks in recruitment and retention.These issues are related to the multifaceted picture of teachers' working conditions, which goes beyond initial recruitment and induction program experiences.The policies described in this section were informed by the LMP paradigm.Many offer strong solutions for recruiting and staffing schools, but have less strategy and insight when it comes to retaining teachers.We know that to tackle this problem we must develop comprehensive solutions rather than recruitment-heavy strategies, using new perspectives and ideas.

Toward a More Comprehensive Picture of Teacher Recruitment and Retention
In addition to SNP, a range of studies exist that offer variations or alternatives to the LMP.The majority of these studies share several characteristics, including consideration of organizational factors, the inclusion of qualitative data, and longitudinal study.The results from these studies highlight the importance of teachers' workplace conditions to teacher retention.These alternatives contribute to a more comprehensive picture of teacher mobility, recruitment, and retention patterns.However, more studies that reveal the complex role of relationships or networks and organizational contexts such as SNP research could help to close the gap.
A well-known study of teacher retention is Ingersoll's (2001) study of the organizational factors affecting teacher attrition.His and similar studies focusing on organizational structure offer a variation on the LMP by focusing on teachers' perspectives on organizational features such as administrative support instead of their individual characteristics (George, George, Gersten & Grosenick, 1995;Singh & Billingsly 1996).These studies indicated that supportive relationships within schools were vital to teacher retention.For example, George et al. (1995) found a significant relationship between teachers' perceived collegial support and their decisions to remain in teaching.
Another oft-cited research project on teacher retention and mobility that incorporated alternative perspectives and used qualitative data was the Project on the Next Generation of Teachers at the Harvard Graduate School of Education (Johnson, Birkeland, Kardos, Kauffman, Liu, & Peske, 2001).One of the findings from the studies was that relationships and the culture of the school comprised a major influence on a teachers' decision to stay or leave.Swars, Myers, Mays, and Lack ( 2009) provide a more recent example of a study that incorporated qualitative data.Their study examined teacher retention at a Professional Development School (PDS).The results were similar to the Next Generation project-the two primary factors related to teacher retention were congruency of beliefs to the organizational norms and the relational needs of teachers.
Finally, longitudinal studies such as Kelly (2004) and the Next Generation project offer more than just a snapshot of teacher mobility; they offer changes in needs and perspectives over time.Kelly's (2004) event history study of teacher attrition revealed the weakness of the link between teacher salary and attrition due to his long view on the data.On the other hand, the climate of a school was a key indicator of attrition.
Although teachers' working conditions were found to be significant predictors of attrition in these aforementioned studies, overall there has been uneven attention to these issues due to the large number and complexity of factors that contribute to these conditions.In their extensive review of research on teacher retention, Borman and Dowling (2008) contend that issues concerning teachers' working conditions need to be examined from alternative research perspectives that focus on social dynamics.Research using the SNP such as social network analysis has the ability to analyze the social aspects of teachers' work in a comprehensive manner because it considers the characteristics of relationships as they relate to the dynamics of the organization.The SNP joins together organizational and individual research level approaches for a multi-level examination of workplace conditions and social dynamics.

What is a Social Network Perspective?
The social network perspective is not new; it is employed regularly in anthropology, sociology, and economics.However, in research on teachers it has only recently begun to appear, as scholars have become more interested in communities of practice and teachers' networks.The perspective is rooted in the concept of social capital.Thus, research from this perspective focuses on the patterns of links and interactions between individuals or groups in a social network and how these trends shape their experiences and choices.The perspective represents a bridge between individual and organizational levels of research.Further, it acknowledges the dynamic nature of networks and relationships.
The most common approach to SNP research is social network analysis, a quantitative approach to characterizing the social ties in a network.Social network analysis can be distinguished from other forms of quantitative research through its focus on the structural characteristics of social networks in relation to attributes of actors in the network rather than a normative focus on attributes or individuals (Wellman, 1983).In this perspective, individuals or groups may be units within a network.This lens changes the types of information and outcomes that are typically found in traditional research.Actors in a network are grouped (classified) after their relationships and interactions have been observed rather than by formal titles or group labels (Haythornthwaite, 1996); this approach brings an understanding of the underlying and informal exchanges that may not be uncovered through traditional quantitative research.
Another SNP approach uses qualitative methodologies to describe and analyze a network or several networks.Typically, research using this type of approach examines the role of trust and the ways in which networks can affect individuals' beliefs, access to resources, and learning (DeVoogt Van Lare & Yoon, 2009).Qualitative methods can also be incorporated into mixed-methods approaches to studying networks (Martinez, Dimitriadis, Rubia, Gomez, & de la Fuente, 2003).Some research on social capital is framed by a perspective that considers social capital to be a community asset, enhanced by fostering connections between community members through common norms and practices (Coleman, 1987;1990;Putnam, 1995).In contrast, social network analysis scholars such as Granovetter (1985Granovetter ( , 2003) ) argue that diverse networks without high levels of interconnectedness (known as "weak ties") offer the highest degree of social capital due to an infusion of new ideas and information from their diverse members.SNP research on knowledge networks and learning communities has found benefits and drawbacks of both.Close-knit, often homogenous communities foster greater degrees of emotional support, but information in the group is often redundant; weakly-linked, heterogeneous networks tend to have more instances of innovation and change but less support from members (Hakkarainen, Palonen, Paavola, & Lehtinen, 2004).
Another element unique to SNP is the dual focus on micro (individual) and macro (structural or organizational) elements of a network (Kilduff & Krackhardt, 1994). 2 The outcomes of SNP research explain the patterns of individual relationships within a larger network, as reflected in Daly, Moolenaar, Bolivar and Burke's (2010) research on teachers and school reform.They studied school reform efforts in San Diego and found that as time went on, populations at each organizational level (teachers, coaches, administrators) coalesced within each group, a development that left each group disconnected and less able to exchange information and resources for the reform efforts.Daly et al. identified the individual characteristics of members of the populations, a process that helped them to determine the reasons for the movement towards group isolation.
The dynamic nature of social networks demonstrated in Daly et al.'s findings has been highlighted in research on knowledge exchange.Several scholars have developed conceptions of networks that consider the dynamics and distributed nature of networks (Hakkarainen, Palonen, Paavola, & Lehtinen, 2004).One of these new dynamic network concepts is the Intensional (sic) Networks (INs).INs are ego-centric, personal networks of people called upon to accomplish a task or project, which may change with each task or need (Nardi, Whittaker, & Schwartz, 2000, 2002).In research on first year teachers' networks of support, Baker-Doyle (in press) identified similar types of dynamic knowledge-exchange networks, known as Intentional Professional Networks and Diverse Professional Allies, which were found to have a powerful effect on early career teachers' ability to locate support and implement curriculum (Baker-Doyle, in press).SNP theoretical findings have reshaped the ways that organizational management is studied and developed.

The Social Network Perspective and Research on the Workplace
For the last 20 years, SNP research has made significant contributions to business scholars' and administrators' understandings of workplace concerns (job performance, promotion, turnover, satisfaction, and management) and organizational practices and behaviors (leadership, innovation, and collaboration).Some of the most frequently researched interpersonal studies are on job satisfaction and performance, and many researchers are concerned with how the balance of interaction and social support affects employees' satisfaction and support.Brass, Galkeiwicz, Greve, and Tsai (2004) note, "[I]solation is probably negatively related to satisfaction, while a higher degree of centrality may lead to interaction with unpleasant others, conflicting expectations, and stress" (p.798).One noted finding in this area is that network factors such as centrality and proximity shape job related perceptions (Ibarra & Andrews, 1993).Another is that centrality and active networking with diverse actors or groups predicts workplace performance (Mehra, Kilduff, & Brass, 2001).A third finding is that people with contacts beyond normal work requirements provide important information and influence in the organization (Brass, 1984).These findings suggest that an understanding and knowledge of the informal networks present in a workplace can help to explain aspects of job experiences and performance.
SNP studies on job mobility have particular relevance to the problem of teacher turnover.More than 15 years before the "revolving door" discussion arose in education circles, Krackhardt and Porter (1986) suggested the same concept as the "snowball effect" in their turnover study (that turnover causes more turnover).Further, they found that turnover occurs in clusters of employees occupying similar roles (structurally equivalent actors) as defined by their perceived communication patterns.More recently SNP scholars have argued that weak ties are a major source of change (including turnover) and that people with stronger ties have a greater impact on turnover rates (McPherson, Popielarz, & Drobnic, 1992;Mehra et al., 2001).
Although there is a wealth of SNP research on workplaces in general, there are far fewer SNP studies on schools as workplaces.Workplace studies offer a glimpse of the type of information we can learn from SNP research on teachers.The following section discusses the factors that make SNP research a valuable approach for considering issues surrounding teacher recruitment and retention and describes key findings from current SNP research on teachers.

Why Social Capital and Social Networks Matter to Teacher Turnover
Research on the social networks of teachers has found that the characteristics of teachers' social networks directly influence their decision to teach or stay in a certain location (Boyd, Lankford, Loeb, & Wycoff, 2005;Thomas, 2007), ability to cope with change (Spillane & Louis, 2002), sense of support (Baker-Doyle, in press;Burns Thomas, 2007), professional learning (Lieberman, 1995), commitment (Riehl & Sipple, 1996), and students achievement (Boyd et al., 2006).All of these factors are important aspects of "workplace satisfaction" and affect teacher attrition rates.Collectively, these findings indicate the important role of relationships and social capital in teachers' professional lives.
Teachers' mobility patterns are closely related to their social network characteristics.Boyd et al. (2006) found that teachers tend to look for positions close to home or in areas similar to their homes.This finding can be explained through the SNP principle of homophily, that people tend to seek out others who are similar to them.In his research on beginning teachers, Thomas (2007) noted that teachers with higher rates of mobility also tended to have a have a more heterogeneous pattern in their social networks, with more ties to persons outside of their own school.This finding is supported by similar mobility patterns in research in non-educational workplaces settings (McPherson et al., 1992;Mehra et al., 2001).One implication from the mobility patterns of these studies is that the more rooted (socially tied) a teacher is in her community, the less likely she is to leave.
The importance of a teachers' network extends beyond the recruitment and selection process.Once in a new school, a teacher's ability to navigate the politics of the schools, cope with reforms and changes, and locate support is also tied to his social network and networking characteristics.Baker-Doyle (in press) studied the social networks of first year teachers and found that when teachers developed "Intentional Professional Networks" (teachers' informal personal networks, often based in the teachers' schools, similar to Intensional Networks), they reported higher degrees of professional support and felt a greater sense of school status than teachers whose support networks were less dense and outside of the school context.In another study on teacher networks, Frank, Zhao, and Borman (2004) found that teachers' social networks were as important in the successful implementation of a new technological reform as access to materials, research, and training.Teachers rely on their networks to interpret and implement new reforms and policies (Coburn, 2001(Coburn, , 2005)).
The teacher turnover problem is not merely about retaining teachers.It is also about retaining high quality teachers.Social networks and networking characteristics have also been identified as important aspects of the development of teacher quality, school capacity, and student achievement.Teacher quality has primarily been conceptualized as a level of human capital (i.e.formalized knowledge) and more recently as disposition.However, a teachers' social capital (the resources and information to which she has access through her network) is also important to her professional learning (Baker-Doyle & Yoon, in press;Frank, Zhao, & Borman, 2004).Teachers' social capital is developed through networking and collaboration practices.As such, schools that emphasize teacher networking have more success in communicating and organizing towards their goals (Penuel, Riel, Krause, & Frank, 2009).In addition, teachers with more social capital tend to have more autonomy in their classrooms (Achinstein, Ogawa, & Speiglman, 2004)-lack of autonomy is one of the major "workplace concerns" that teachers' cite for leaving the profession (Ingersoll, 2001).Finally, the capacity of schools to be successful in raising student achievement is increased when teachers are encouraged to collaborate and develop networks (Spillane & Louis, 2002).
A comprehensive examination of teachers' workplace experiences must also consider longterm commitment.SNP research has revealed that the characteristics of teachers' social networks may be an indicator of their long-term commitment.Riehl and Sipple (1996) examined secondary school teacher commitment and found that the level of instructional support teachers reported that they received from peers predicted their commitment levels.Further, Thomas (2007) found that early career teachers with stronger ties within their schools were more likely to remain in those schools.Social networks play a considerable role in shaping teachers' lives and choices throughout their careers, from identifying schools in which they want to work, to seeking out support and information, to becoming effective teachers, and, ultimately these networks affect teachers' commitment to the profession.

Using SNP to Examine School Policies: A Case Example
While LMP highlights the attractiveness of an approach (Guarino, Sanibanez, & Daley, 2006), it does not provide a strong conceptual framework that can reveal the complex nature of social interactions and their affect on teacher attrition.Moreover, the social dynamics of teacher quality and professionalism are challenging to study in an LMP model.The SNP can help to explain the functioning of policies and approaches in a way that sheds greater light on the social aspects of teaching.One school reform policy that has its roots firmly in the LMP is reconstitution.In 1997, the Chicago Public Schools reconstituted several low-performing schools (Malen, Croninger, Muncy, & Redmond-Jones, 2002;Rice & Malen, 2003).Although the theory of action is based in notions of human capital and seemed sound, it was one-sided; it lacked consideration of the role of relationships in teacher retention, to great detriment.
Reconstitution is conceptualized as a method for increasing the human capital in school through re-staffing.In Chicago teachers in selected schools were required to reapply for their positions and administrators were directed to hire only those with the best qualifications.From a very basic LMP standpoint, Malen et al. (2002) described, such "reconstitution reforms either assume that large pools of talented and dedicated people are awaiting the opportunity to work in troubled schools or that such pools can be generated by the appeal of a 'fresh start'" (p.127).However, this theory of action did not hold up when tested in Chicago (Allensworth, Ponisciak & Mazzeo, 2009;Hess, 2003;Malen, Croninger, Muncy ,& Redmond-Jones, 2002).The social complexities of the workplace had a major impact on the outcomes.
Rice and colleagues went beyond LMP in their research and incorporated an SNP approach into their study design (Rice & Croninger, 2005;Rice & Malen, 2003).In addition to considering the effects of the policy on financial and human capital, they considered social capital, conceptualized as trust and commitment within the teacher community.They found that the act of reconstitution destroyed the social networks in the school, a cosnquence that affected teacher morale and attraction to the school.One teacher noted, "We were a close-knit family.We helped each other.When they broke up the family, things went downhill at this school, and it has gone downhill since."(Rice & Malen, 2003, p.652).
Several aspects of the schools "went downhill."First, the schools had higher rates of turnover for several years, causing mismatches in staffing and lower quality teaching staff.In a similar study of the Chicago reconstitution effort, Hess (2003) noted, "[S]ome reconstituted schools ended up hiring their counterparts' rejects.Other schools limped along… plugging gaps with substitute teachers" (p.308).As of 2009, more than 10 years later, the Chicago School District continues to have challenges in recruiting and retaining teachers (Allensworth, Ponisciak, & Mazzeo, 2009).
Second, there was a high degree of distrust among the reconstituted staff (Hess 2003) and teachers were less likely to collaborate; they told researchers that they "pulled back" from the school community and tended to be wary of working with others (Rice and Malen, 2003, p. 653).The school capacity declined, and student academic progress did not improve significantly.Hess ' (2003) longitudinal study of the effects of the reconstitution reported that the students' slightly higher test scores should be attributed to a change in student population rather than staffing changes, and he concluded, "reconstitution in Chicago turned out to be a one-time event which was not particularly successful, either in creating significant change in the...schools or in occasioning significant improvement in student achievement" (p.326).
While the LMP research paradigm justified the theory of action, the SNP approach considered the role of social interactions and revealed the complex social factors that contributed to its failure.As reconstitution continues to grow in popularity as a strategy to improve teacher quality, it will be necessary to incorporate alternative perspectives such as the SNP into research and policy paradigms to reveal potential problems in assumptions and theories of action.

Conclusion
I have argued that the labor market perspective (LMP) is limited in its ability to uncover and conceptualize the roles of social capital and social networks in the process of teacher recruitment, hiring, and retention.Further, these limits have left holes in our knowledge about effective recruitment-retention practices, focusing mainly on recruitment rather than on comprehensive approaches to hiring and keeping high quality teachers or helping teachers learn to create their own support networks.Thus, with the exception of a few promising programs, many district staffing efforts are recruitment-heavy or conducted in a piecemeal fashion.The SNP approach provides new questions, insights, and theories for understanding the ways teacher recruitment and retention policies function and affect teacher quality and attrition.It is evident from this review that social capital and social networks are a significant factor in teacher recruitment and retention and require additional research.
Researchers and policy-makers should take a more comprehensive approach to studying, designing, and implementing recruitment-retention efforts; we need to look beyond the traditional LMP approach and strengthen our focus on teachers' professional communities and relationships.A combined research approach that integrates LMP concepts with SNP would likely provide greater insight and understanding about such programs.An example of one project that is using the SNP to examine the new teacher experience is the Michigan Indiana Early Career Teacher (MIECT) research project.Early findings from this study have already revealed patterns relating networking characteristics and commitment levels and hidden patterns of mentoring for early career teachers (Low, 2009;Youngs, Frank, & Pogodzinsky, 2009).
Even as the recession of 2009 propelled more individuals to seek teaching jobs, the problem of retaining high quality teachers beyond their first five years remains an urgent issue.The labor market perspective produces a wealth of knowledge about attrition trends and the attractiveness of certain schools to various populations of teachers.However, what it lacks-the ability to describe and analyze complex social factors-is a key element of fostering retention and sustainability.A social network perspective can help to address this missing link and, therefore, can help scholars and policy-makers re-conceptualize attempts to solve the problem of teacher turnover.