THE VALUE OF TEACHER QUALITY For most it comes as no surprise that research has consistently shown that teachers can have profound eff ects on their students. Research dating all the way back to the “Coleman Report” (Coleman, 1966) found teacher quality to be the most important schooling factor infl uencing student achievement. Recent fi ndings, based on better data and more rigorous statistical methods, confi rm this conclusion and furthermore show that there is considerable variation in quality among teachers (Aaronson et al., 2007; Goldhaber et al., 1999; Rivkin et al., 2005). 1 Hanushek (1992), for instance, fi nds that the quality of a teacher can make the diff erence of a full year’s learning growth, and Chetty et al. (2013) fi nd that diff erences in teacher quality explain a variety of later life outcomes, such as college attendance and labor market earnings. 2

While it is common to discuss teacher quality in abstract terms, the nuts-and-bolts definition of the concept is to a large extent in the eye of the beholder. One notable change in teacher research and policy over the last decade is the shift toward assessing teacher quality based on teachers’ “value added,” that is, their contributions to student achievement on standardized tests. 3 Indeed the research cited above referring to consequential variation in teacher quality is based on value added. The use of value added to assess individual teachers is controversial for a variety of reasons (discussed more fully below). Probably the most fundamental reason is that student achievement on tests captures only one aspect of what teachers contribute to students. Teachers no doubt contribute in many different ways to students’ intellectual and emotional growth, implying that their quality should be measured along a number of dimensions. Historical measures of teacher quality have generally been associated with the set of credentials that a teacher holds. 4 It is also commonplace to associate the quality of teachers with their classroom practices (e.g., Bell et al., 2012; Hill et al., 2011 and 2012). Nonetheless, in this chapter I primarily focus on research that uses gains in student test scores as a metric for teacher quality. I do this not only because test scores are arguably the best indicator of teacher performance (at least for the relatively small proportion of teachers for whom they can be calculated), but

also for a more practical reason: a good deal of education research focuses on test scores as an outcome because there is little variation in other measures of teacher performance (Weisburg et al., 2009). Hence, throughout the chapter, I use the term “teacher quality” interchangeably with “teacher effectiveness,” “teacher performance,” and “value added.” For a richer discussion of teacher effectiveness, see Donaldson and Papay in this volume.