graciously sponsored the survey, and Briggs McAndrews at NYSCOSS provided significant help with all aspects of the survey. Deborah Cunningham, Charles Szuberla, Johanna Duncan-Poitier, Nancy Willie-Schiff, and Joe Porter of the New York State Education Department (SED) reviewed the survey and provided helpful comments. Brenda Budka at SED provided the teacher certification data, and patiently answered our many questions about this data. Dan Goldhaber and Michael DeArmond of the University of Washington had a number of valuable suggestions on both the content and design of the survey. Willow Jacobson at the University of North Carolina, Pat Ingraham at Syracuse University, and Mark Robbins and Bill Simonsen, at the University of Connecticut, critiqued the survey instrument, and proposed a number of revisions to reduce length and improve clarity. We also want to thank Jeff McLellan, Kim Kirsch, and Bob Bifulco for their comments on preliminary draft of the report. We appreciate the willingness of Robert Strauss of Carnegie Mellon University to allow us to borrow from his survey on teacher hiring practices in Pennsylvania. Errors and omissions are, of course, solely the responsibility of the authors.

On behalf of:

the New York State Board of Regents, and education scholars, researchers and policymakers from New York and around the country. By bringing education researchers together with policymakers, the Consortium seeks to address critical questions about the state of public education in New York. The Consortium’s mission is to foster an exchange of information and ideas that informs education policy decisions, assist education policy decision-makers in identifying issues that require initial or further study, and produce quality and timely research that contributes to the development of sound education policy. The Consortium is housed in the Center for Policy Research, Rockefeller College of Public Affairs & Policy, University at Albany – SUNY. Condition reports are written using education data compiled by New York State agencies and from data independently collected by the researchers. Consortium researchers work closely with agency staff to identify and analyze trends in school district spending, school staffing and student performance. The condition reports describe various characteristics of New York State's public school systems and also identify potential areas for further study.

Acknowledgements: This project would not have been possible without significant help from a number of

Abstract not found

Many school districts experience difficulties attracting and retaining teachers, and the impending retirement of a substantial fraction of public school teachers raises the specter of severe shortages in some public schools. Schools in urban areas serving economically disadvantaged and minority students appear particularly vulnerable. This paper investigates those factors that affect the probabilities that teachers switch schools or exit the public schools entirely. The results indicate that teacher mobility is much more strongly related to characteristics of the students, particularly race and achievement, than to salary, although salary exerts a modest impact once compensating differentials are taken into account.

Funding for this study is generously provided by the Joyce Foundation. Silicon Valley-based SRI International is one of the world’s leading independent research and technology development organizations. Founded in 1946 as Stanford Research Institute, SRI has been meeting the strategic needs of global markets for more than 60 years. SRI’s Center for Education Policy studies reforms that hold promise for improving the K-16 system of schooling and lifelong learning. The Center conducts research and evaluations on the design, implementation, and impact of educational programs, especially improvement efforts targeted at disadvantaged students.

There is no longer any argument about whether reducing class size in the primary grades increases student achiev ement. The research evidence is quite clear: It does. (Molnar, 1999: 53) However, despite this clear statement, when considering the implementation of smaller