Skip to Content

  • Policy Report
  • An Overview of Research on the Effectiveness of Retention on Student Achievement for New York City School Children

    This paper analyzes 25 years of research documenting the failure of single-test retention policies, looking in particular depth at New York City’s past failed retention policies and the current data on Chicago’s retention policy. The paper was accompanied by a sign-on letter in opposition to the Mayor’s announced policy to hold back fifth graders on the sole basis of their scores on standardized tests.

    Sep 27, 2004

    Yellow pencils on top of a black and white composition notebook. (Photo by Kelly Sikkema on Unsplash)
    Photo by Kelly Sikkema on Unsplash

    The letter was signed by over 25 academics, heads of organizations, and experts on testing, as well as educators and advocates. If passed, the Mayor’s proposed policy may mean the retention of approximately 13,000 fifth graders next fall.

    The protest letter, signed by two past presidents of the American Education Research Association, the nation’s premier organization of educational researchers, the past president of the National Council on Measurement in Education, as well as two heads of educational research think tanks, heads of educational advocacy groups, educators and parents, makes clear the widespread opposition to the Mayor’s retention proposal.

    It is puzzling why the Mayor would want to implement a policy that has failed New York’s schoolchildren in the past, comes with a tremendous price tag, and is not supported by over a quarter of century of research. There are clear solutions, such as academic intervention and class size reduction that do work. This is where our educational funds need to go, not on a policy destined to fail.”

    Jill Chaifetz, Executive Director of Advocates for Children

    Leonie Haimson of Class Size Matters stated, “There is overwhelming agreement among academics, experts on testing, and educators that retention hurts rather than helps students and leads to significantly higher drop out rates. Imagine that the Mayor decided, on his own, that a certain surgical procedure should be used in all the public hospitals in the city, even though the professional consensus was clear that the procedure would lead to much higher rates of complication and mortality. Would he be able to impose his views on the practice of medicine? I think not. So why should it be any different in the field of education?”

    As many of the researchers who have signed on have pointed out in their research, assessing a child’s actual level of achievement solely on the basis of one test is inherently unreliable, given the large statistical margin of error and the inherent variability of student performance. Even the two companies that produce the 3rd grade standardized exams, Harcourt and CTB-McGraw, are on record that the decision to retain a child should never be made on the basis of test scores alone.

    Moreover, in the 1980s, New York City attempted a similar policy of large-scale retention during its “Gates” program, which failed miserably to improve student achievement. As the Department of Education stated in its own Resolution to rescind the Gates program in 1991:

    It has been determined that Promotional Gates had little positive impact on students. Each year more than one-third of the students who were held over and attended Gates classes still failed to meet promotional standards… A longitudinal study indicated that a disproportionate percentage of students held over in Gates classes became dropouts… There is no evidence, therefore, that holdovers make academic progress, although there is evidence that holdovers demonstrate greater social and emotional difficulties.

    Those signing on to this letter today want to prevent history from repeating itself. The research is overwhelming that holding back low-achieving children harms rather than helps their educational prospects, and instead leads to higher dropout rates. In addition, if this policy is implemented it will also likely have a disproportionate and damaging effect on poor and minority schoolchildren, further diminishing their chance of success.

    Related Policy Resources