Teachers make a difference. Everyone—students, parents, teachers, principals, and even casual observers—has always known this. But how much of a difference teachers make has only recently been estimated. Teachers are arguably the most important school factor affecting student achievement;¹ but, especially important for the purposes of this paper, the variation in impact across teachers is striking.² Indeed, according to one estimate, the achievement gap between advantaged and disadvantaged students could be closed if disadvantaged students had a good teacher for just five years in a row.³

Education reformers have seized on findings about teacher effects. It was good news that could be acted on through public policy. Whether it is concerns about global competitiveness or the tremendous disparities in educational achievement within the United States, public schools are under intense pressure to improve. Research findings suggest that one solution is clear and, at least on the face of it, something policymakers can tackle. We know it is possible for teachers to make large differences in student learning. Policies and practices affecting where and how teachers work must then reflect the importance of teachers in the educational process. The task is to develop policies that increase the number and equitable distribution of teachers who promote learning at high rates.

Some reformers focus on improving skills through better training, both pre-service and in-service;⁴ others focus on teacher recruitment strategies.⁵ But the reform that has generated the most attention, and the most heat at the local, state, and national level, is giving teachers performance incentives. There are two theories of action that undergird this approach. One is that incentives matter and teachers would produce greater student learning (or at least put greater effort into producing student learning) if there were incentives or rewards for doing so. The second is that because current compensation policies generally do not significantly differentiate among teachers based on characteristics other than longevity of service and courses taken, introducing compensation-based ways of recognizing and rewarding performance could make teaching a more attractive career option. Both approaches also assume that rewarding performance might also put to better use funds now in the traditional compensation system, which rewards teachers for characteristics (such as certification and experience) that are at best weakly related to teacher effectiveness.

Almost every area of policy—from tax policy to health and environmental policy—uses incentives of some form to promote desired ends. In education, they are strongly resisted. One reason for resistance is simply that accountability based on performance is relatively new to education. Traditionally, educational accountability was based on inputs: counts of students, teachers, class size, etc. School level accountability began to emerge in a few states in the late 1980s and 1990s; nationally, however, No Child Left Behind (NCLB) established school level accountability only in 2001. According to the law, schools that persistently do not perform face consequences. Not surprisingly, these policies met with considerable push-back from the education establishment when they were introduced, and they continue to face substantial opposition today.

When performance accountability is focused on teachers, the issues become magnified and the dynamics intensify. Teacher-based performance incentives fundamentally change the name of the game for teachers: measured student achievement gains become the focal point of human resource management strategies. Doing this in a way that is effective, fair to teachers, and defensible is a complicated task, as we discuss later. But teachers have a powerful guardian to protect their interests: teachers unions.⁶

Teachers unions gained purchase in American education in the 1960s largely by piggybacking on the precepts of industrial unions, focusing principally on protecting their members from unreasonable management demands and emphasizing "bread and butter issues." Today, in many states, collective bargaining between unions and management largely determines the rules governing the education workplace—schools. Union resistance to differentiation of its members, especially on the basis of performance, is consistent with bargaining for the collective. But growing demands for accountability and improved performance in education, along with large new databases that link tested student performance to individual teachers, are presenting challenges to the unions' positions because ways to measure performance that are defensible, effective, and fair are realistic possibilities in public policy.

The purpose of this paper is to examine the interplay between the emerging policy focus on teacher performance incentives and the response of teachers unions. We focus first on the policy shift itself. Without an understanding of the nature of the teacher's job and its tradition, it is difficult to understand the tug-of-war both for and against performance incentives. We then examine the extent to which, and the ways in which, teachers unions are responding to demands for increased accountability and the possible consequences of this relationship for shaping policy and practice.

Read the full paper (PDF) on Vanderbilt's "Performance Incentives" Conference Web site. 

Introduction Endnotes

^1. Increasing teacher quality easily trumps reducing class size as an educational investment (Hanushek, 1992).

^2. A large number of studies substantiate this observation, e.g., Sanders and Rivers (1996); Ballou et al. (2004); Hanushek et al. (2004); Kane et al. (2006); Rockoff (2004).

^3. Rivkin et al. (2005)

^4. See, for example, Hill (2007) for review of research on the effects of different forms of professional development on teacher effectiveness.

^5. Boyd, Lankford, Loeb, Rockoff and Wyckoff (2007), for example, show the likely student achievement effect of hiring teachers with different qualifications and assigning them to high poverty schools is reducing the achievement gap. The teachers were mainly teachers from the New Teacher Project and Teach for America who had, for example, higher SAT and other test scores. Research has shown, however, that traditional teacher characteristics that affect pay, such as certification, years of experience (beyond the first few years), and holding a master’s degree, have little relationship to teacher effectiveness. See, for example, Kane, Rockoff and Staiger (2006).

^6. Interestingly, when observed across states with mandatory and non-mandatory collective bargaining statutes, these laws are seen to have only modest effects on negotiated contracts and, in turn, on teacher practice in the classroom. For a discussion of these variations, see Hess and Kelly (2006).