If you haven't read it already, I highly recommend Washington Post reporter Bill Turque's analysis of Michelle Rhee's legacy one year after she left the D.C. public schools. Turque writes about the "churn and burn" in the D.C. teacher corps since the introduction of the controversial new IMPACT teacher evaluation and merit pay system: One-third of all teachers on the payroll in September 2007 no longer work for the district, and inexperienced teachers are more clustered than ever in low-income schools and neighborhoods. We know this is problematic because DC's own data shows that 22 percent of teachers with six to 10 years of experience are rated "highly effective," compared to just 12 percent of teachers with less than six years experience.
That said, if inexperienced but minimally-competent teachers are replacing grossly incompetent ones, the high turnover might be a better option than leaving bad teachers in the classroom. In the case of D.C., is that what's happening? Democrats for Education Reform has a new report on IMPACT with some helpful charts that put the "one-third attrition" stat in context. First, let's take a look at how IMPACT evaluated DC's 4,000 teachers' union members during the 2010-2011 school year.
The small gray wedge represents the 2 percent of teachers who were rated "ineffective" and immediately terminated. Another 2 percent were rated "minimally effective" two years in a row and then terminated. These two groups account for about 200 teachers. The four-year turnover rate of one-third, therefore, is vastly larger than the 4 percent of "ineffective" teachers in the district. Even if every single teacher rated "minimally effective" for just one year had quit, these three problematic categories would account for only 17 percent of the teacher corps--about half the actual four-year attrition rate.
The comparison isn't perfect because these IMPACT figures are a snapshot from a single school year and, as Turque has reported, D.C. does not release its annual teacher attrition rate--the number of teachers who don't come back to work from spring to the following fall. But extrapolating from the available data, it seems clear that the majority of teachers leaving D.C. are currently "effective," or at least have the potential to become effective teachers over time. Indeed, the DFER report demonstrates that D.C. teachers rated "minimally effective" who stay in the district have a decent chance of improving under IMPACT.
I think the longterm question on IMPACT and other new evaluation plans is whether promising teachers choose to stick around to work under these systems. Of course, some teacher attrition is inevitable, since teachers retire, move away, and go on maternity leave. But the average yearly teacher attrition rate in a low-poverty American public school is 12.9 percent, compared to 20 percent in a high-poverty school. Since unwanted teacher attrition costs the typical urban school district tens of millions of dollars annually--and disporportionately affects low-income kids--a good test of any teacher quality reform is whether it improves the retention of effective teachers, not just whether it results in the firing of ineffective ones.
The evidence from DC is mixed, since many more teachers are leaving the District than have been deemed ineffective or unlikely to improve. At the same time, those who choose to stay despite the challenges and tumult seem to be gaining some professional development benefits.
The DFER report contains testimonies from teachers who felt IMPACT improved their practice
Jay Matthews' column on one well-regarded DC teacher who found the system unhelpful
Jay Matthews' column on a DC princpal who dislikes IMPACT