Association for Public Policy Analysis & Management

Despite widespread concern about teacher shortages, many schools receive more applicants for open teaching positions than they have vacancies and many newly-certified teachers do not get hired into teaching positions at all (Cowan, Goldhaber, Hayes, & Theobald, 2016; Jacob, 2007). This suggests that many administrators have meaningful discretion when hiring new teachers, and given that teachers vary considerably in their effectiveness (e.g., Chetty, Friedman, & Rockoff, 2014; Rivkin, Hanushek, & Kain, 2005), how this discretion is exercised may have important implications for students and schools. Yet little is known about how school systems can effectively hire teachers. An extensive literature documents school principals’ preferences and priorities when hiring teachers (e.g., Cain-Caston, 1999; Harris, Rutledge, Ingle, & Thompson, 2010), but often relies on administrators’ self-reports, which may not accurately reflect how teachers are hired in practice, and there is evidence that actual hiring is often rushed and “information-poor” (Levin & Quinn, 2003; Liu & Johnson, 2006). It is also not clear how principals should hire teachers given that teacher effectiveness has often proven difficult to predict using easily-observable teacher characteristics (e.g., Harris & Rutledge, 2010; Rockoff, Jacob, Kane, & Staiger, 2010). This may make detailed teacher screening attractive, though only a few studies examine such screening directly (Goldhaber, Grout, & Huntington-Klein, 2017; Jacob, Rockoff, Taylor, Lindy, & Rosen, 2016).

We contribute to this literature using data from a new screening system recently adopted by the Los Angeles Unified School District (LAUSD). In LAUSD prospective teachers apply to the district office, where they are screened using a highly standardized system with eight components (e.g., a writing sample and the delivery of a sample lesson), each scored according to rubrics aligned to district goals (e.g., employee evaluation criteria). We observe all individuals who meet the minimum passing requirements of screening and are thus deemed eligible to be hired by school administrators, and we link those who are hired to detailed teacher- and student-level demographic, evaluation and test score data. These data thus capture many applicant characteristics that are often difficult to observe and allow for analyses both of applicants’ relative employment prospects and of the predictive validity of screening instruments.

Preliminary results suggest that, controlling for certification and the time at which they enter the hiring pool, prospective teachers’ communication skills and professional references are predictive of their probability of being hired, but measures of academic ability and subject area preparation are not despite being predictive of effectiveness. Though newly-hired teachers in general perform below the district average as measured by both classroom observations and student achievement, their screening scores are meaningfully predictive of their success once hired. New teachers who receive higher screening scores have substantially lower probabilities of receiving unsatisfactory evaluations and take fewer discretionary absences, and their students experience larger test score gains in both math and English/language arts. In addition, we find suggestive evidence across time and between districts that the shift to the new screening system improved hiring outcomes in LAUSD relative to other similar districts and schools.