Response by Dr. Alan Ginsburg to Paul Peterson's April 11, 2011 article in Education Next and his Washington Times story taking issue with my analysis showing that Rhee's DC gains were no better than those of her two predecessors Vance and Janey. (April 15, 2011)

“In the case against Michelle Rhee,” Paul Peterson presents three alleged flaws in my analysis “ The Rhee DC Record: No Better Than Her Predecessors Vance and Janey.” Two of Peterson’s proposed flaws simply misrepresent my methodology. The third makes the highly questionable adjustment that the DC school system deserves credit only for DC gains above the national average NAEP gain, with Peterson giving credit to some unknown cause for DC gains up to the national average.

First, Peterson claims that my analyses did not adjust for the fact that “Rhee was in office for only two years, while Vance was in office for three and Janey for four.” This is a misrepresentation of my methodology. The analyses (Exhibit IIB and IID below) in the report clearly represent the average annual NAEP score gains. In math, Rhee’s annualized gains fall between Vance and Janey and in reading the annualized gains are about equal between Janey and Rhee.

Second, Peterson also claims that I used the DC NAEP sample that “in 2009 did not include students attending charter schools… while in 2007 all charter school students were included.” This is also a misrepresentation. My report clearly specifies that I chose the state rather than the urban NAEP series precisely because of its consistent treatment of students in charter schools over the full 2000-2009 period.

Third, and crucial to Peterson’s claims, is the assertion that the DC NAEP score improvement should be computed only as the excess above the national average NAEP gain. In the highly decentralized U.S. education system in which the federal government is prohibited from specifying curriculum, removing the national average NAEP gain because it reflects some national cause is unwarranted.

Furthermore, the reason that removing the national average NAEP score gain works to Rhee’s favor is that during Rhee’s tenure NAEP gains were substantially lower than during the earlier 2000-07 period. Peterson offers no explanation for what occurred nationally that warrants giving Rhee more credit for her NAEP gains than Vance or Janey. One might cite the fiscal crisis as a major national event, but DC, with its stable employment, was less affected by the crisis than any region of the country. In fact, its per pupil spending between 2007-09 jumped by nearly 20 percent over two years.1

Conversely, although DC math gains over the full 2000-07 period were higher than any state and at grade 4 and DC gains were tied for fourth highest at grade 8 , Peterson would not give DC much credit for these outstanding gains during this period. Prior to Rhee’s tenure.

One valid challenge in estimating DC gains for which there are no available answers is how to treat gains under Rhee between public school students attending charters and non-charter public schools. Although charter and non-charter gains during Rhee can be separated under NAEP, the Rhee administration closed a number of schools in DC during the 2007-09 period causing enrollment in charters to increase by half and enrollment in non-charter DC public schools to decrease by one quarter in only two years. With such large demographic shifts, the National Academy report cited in Peterson’s paper is correct in stating that the longitudinal tracking of students is essential to estimating DC gains. Unfortunately, the DC contract for longitudinal data system development had to be cancelled for malfeasance in 2009.

What is clear is that DC score gains for all public school students have been impressive over the whole 2000-09 period. That includes the public school student gains under Vance and Janey that were comparable to those under Rhee.


[1] Mary Levy March 14, 2011 testimony before the mayor of the District of Columbia hearing on the DC budget for 2012.

Title: The Rhee DC Record: Math and Reading Gains No Better Than Her Predecessors Vance and Janey

Author: Dr. Alan Ginsburg, Director of Policy and Program Studies, U.S. Dept. of Education (Retired)

Feb. 11, 2011

Former DC Chancellor of Schools Michelle Rhee has gained national attention for her tough but controversial reform policies focused on large-scale firing of DC teachers. Her teacher policies have been widely cited as a lever that enabled Rhee to produce an unprecedented turnaround in low DC students’ scores. Notably, Rhee has been featured on the cover of Time magazine with a broom symbolic of sweeping out ineffective teachers and a movie, Waiting for Superman, has been made that included her perceived historic DC success. Based on these perceived extraordinary DC test score gains, she is spearheading a national effort StudentsFirst to raise a billion dollars to replicate nationally her DC policies that include references to her "unprecedented growth" in DC student outcomes.

Experts are divided over the merits of Rhee’s tough teacher policies. Critics have cited their potential to weaken teacher morale and that “teachers deserve respect, not condemnation” (McGuire, 2010). Others see teacher firings and polarization as a necessary price for student progress (Rotherham, 2010). Therefore, it is important to evaluate the effectiveness of Rhee’s policies to determine just how much better DC students progressed under Rhee than under her predecessors.

This paper examines DC’s mathematics and reading scores on the federal government's independent National Assessment of Education Progress (NAEP) since 2000 for all students and by race / ethnicity. It concludes that Rhee did not initiate the DC schools’ test-score turnaround when she took office in 2007. DC’s NAEP scores had already steadily improved under her two predecessors, Superintendents Paul Vance and Clifford Janey. Moreover, the rates of DC score gains under Rhee were typically no better than the rates achieved under Vance and Janey. The paper calls for a careful evaluation of the long-term consequences of the large-scale firing of teachers before launching a nationwide initiative and suggests examining the effectiveness of policies to improve teacher quality along with teacher accountability measures.

The paper is available at The Rhee DC Math And Reading Record .pdf.