This post is from John Krieg (Western Washington University), Roddy Theobald (American Institutes for Research) and Dan Goldhaber (University of Washington Bothell)
A considerable body of evidence suggests that the market for educators is “local” in the sense that teachers are more likely than other professionals to be hired into schools near their original hometowns. Evidence from New York State shows teachers are likely to begin their teaching careers near their homes a phenomenon referred to as the “draw of home.” This draw of home appears to be a combination of preferences of prospective teacher candidates to apply to and accept jobs from districts near their homes as well as a negative preference that districts show towards hiring applicants originating from far away.
The draw of home has potential consequences for the distribution of teachers across districts. Using data from Washington State, the setting for our study, we find that over the course of a decade, three of the four largest teacher education programs produced teachers who came from 177 of the state’s 295 school districts. In other words, 40% of the State’s school districts did not have one of their former students graduate from one of the larger teacher education programs over the course of 10 years. This is more troubling given that college-bound students tend to originate from advantaged school districts. If the draw of home is strong, then these advantaged districts will tend to have first choice among future job applicants leaving districts that send fewer students into schools of education at a hiring disadvantage. Our research confirms a strong draw of home in Washington State. After controlling for district and individual observables, newly minted Washington teachers are 10 times more likely to start teaching in their hometown district than they are a district 50 miles away from their hometown.
Given the apparent strength of the draw of home, we investigate a potential mechanism to alleviate the discrepancies in teacher supply faced by districts. One unknown component of selecting into jobs is the impact that student teaching has on initial job placement. In the State of Washington, 15% of first teaching jobs occur in the building in which student teaching took place, 40% of first jobs are found in the district that hosted the student teacher. This raises the possibility that student teaching placements might be used strategically to overcome the disadvantage faced by districts that send few of their students into schools of education.
Using data from 6,023 hired student teachers, we find that student teaching placements are a much stronger predictor of where an individual teaches than their hometown. For instance, even after controlling for their hometown location, new teachers are 80 times more likely to take a job in the district where they student taught than a district 50 miles away. This raises the possibility that schools of education can use their student teaching placements to direct future job applicants to districts that struggle hiring teachers. One piece of evidence that suggests this might be an appropriate policy is that novice teachers generate higher value added scores when they teach in buildings similar in demographic characteristics to where they student taught. Thus, purposefully placing student teachers in needy districts may have the double benefit of increasing job placement rates of student teachers who are especially prepared to succeed in these districts.
The full study is in John M. Krieg, Roddy Theobald and Dan Goldhaber, “A Foot in the Door: Exploring the Role of Student Teaching Assignments in Teachers’ Initial Job Placements,” Educational Evaluation and Policy Analysis, June 2016, vol. 38 no. 2 364-388.