likely to increase in the coming years as retirements increase (for reasons described below) providing more opportunities for the remaining experienced teachers to transfer from low-performing schools. Indeed BOE has failed to employ sufficient certified staff in its very worst schools, known as schools under registration review (SURR), 19 despite a mandate from SED that it hire only certified teachers in such schools as of September 1, 1999. The Regents have mandated that all schools shall have only certified teachers by 2003. The evidence suggests that BOE will have great difficulty fulfilling this mandate. Special education programs have a high proportion of uncertified teachers. In District 75, which comprises all of the self-contained special education schools for severely and profoundly disabled children, 25% of the teachers are uncertified. In bilingual special education classes, nearly 50% of the teachers are uncertified. The Regents' decisions to remove uncertified teachers from SURR schools, and from the system entirely by 2003, are evidence that uncertified teachers are less able to provide New York City public school children with a sound basic education. 2 . Pa s s a ge Rat e s o n C e r t i fic at i o n E x a m i n at i o n s The probative evidence at trial demonstrates that passage rates on certification examinations are predictive of teacher [187 [187 Misc.2d Misc.2d 28] 28] performance. Such evidence included expert testimony by Professor Ronald Ferguson of the Kennedy School of Government at Harvard University. Using voluminous data collected in Texas during the 1980's, Dr. Ferguson convincingly demonstrated a strong positive relationship between teacher quality as measured by certification test scores and student achievement. Defendants' expert, Professor Eric Hanushek of the Economics Department at the University of Rochester, drew an opposite conclusion from Texas data concerning teacher test scores. Apparently this data was from more recent years than those studied by Dr. Ferguson, but Dr. Hanushek's testimony on this matter was terse. Dr. Hanushek testified that data from Texas tends to show teachers with higher test scores do not affect student outcomes. As Dr. Hanushek set forth only his conclusion regarding the data without any discussion of his methodology, his testimony does not effectively rebut Dr. Ferguson's. Plaintiffs' expert Dr. Langford demonstrated that New York City public school teachers have a much higher failure rate on the State's certification examinations than do public school teachers in the rest of the State. For teachers employed in 1997-1998, the first-time failure rate of the New York City public school teaching force on the basic liberal arts and science test (LAST), which is now required for all teachers, was 31.1% compared with 4.7% for teachers elsewhere in the State. The average first score on the LAST for City teachers was 236.3 (220 is a passing score) while the average first score on the test for teachers in the rest of the State was 261.6. New York City teachers employed as of 1997-1998 also did poorly on their first attempts on the elementary and secondary assessment of teaching skills written tests (Elementary ATS and Secondary ATS). Nearly 27% of New York City teachers failed the Elementary ATS, compared with 3% of their peers in the rest of the State; 25.7% failed the Secondary ATS, compared with 3.5% in the rest of the State. On many of the content or subject matter examinations, which test teachers' knowledge of their particular subject, New York City teachers' failure rates were even higher. For example 42.4% of the math teachers currently teaching in New York City's public schools failed the math content examination at least once. Dr. Langford's testimony was undercut somewhat by one of defendants' experts, Professor Michael Podgursky, Chair of the Economics Department at the University of Missouri. Dr. [187 [187 Misc.2d Misc.2d 29] 29] Podgursky testified that Langford's comparison of City teachers' test scores with those of teachers in other parts of the State may have been marred by incomplete data. In particular, Dr. Langford's data appeared to have a disproportionate number of test scores for New York City compared with other areas in the State. Plaintiffs opine that this may have been a result of the admittedly higher amount of turnover in New York City schools, which would result in more teachers taking the tests in the City, but this was not established as a fact. Additionally, because of the unavailability of test scores for teachers who were hired prior to the mid-1980's, Dr. Langford's test score analysis covers about 40% of New York City's teachers. Despite these valid criticisms of his study, Dr. Langford's testimony concerning teacher test scores is some evidence that New York City teachers are in general less qualified than those in the rest of the State. 3. Experience The unrebutted evidence validates the unremarkable proposition that teachers, like any professionals, frequently require several years' experience to achieve competency. The court finds that teaching experience of two years or less is correlated with poor teacher quality. New York City public school teachers tend to have fewer years' experience than teachers in the remainder of the State. The attrition rate of new teachers is more than 50% in their first six years according to BOE data. In addition to negatively affecting teacher quality, high turnover can negatively affect a school's cohesiveness and ability to create cooperative effort among teachers. The large number of inexperienced teachers— who, like uncertified teachers, are disproportionately assigned to the schools with the greatest number of at-risk students— makes it more difficult for New York City public schools to meet the needs of their students. 4 . C o l l e ge o r U n i v e r s i t y A t t e n d e d a n d D e g r e e O b t a i n e d Finally, plaintiffs presented probative evidence that the average New York City public school teacher attended a less competitive college than the average public school teacher in the rest of the State. For this analysis, plaintiffs ranked colleges using: (1) Barron's college rankings, (2) the average SAT scores and grades of admitted high school seniors, and (3) the average scores on State certification exams of the colleges' graduates. According to Dr. Langford's findings, which are based on State [187 [187 Misc.2d Misc.2d 30] 30]