A 2004 report linked on the NEA’s teacher diversity website lists the evidence that supports that teacher diversity does indeed impact on closing the achievement gap.
In particular, they highlight that
- Students of color tend to have higher academic, personal, and social performance when taught by teachers from their own ethnic groups, [and]
- Teachers from different ethnic groups have demonstrated that when students of color are taught with culturally responsive techniques and with contents-pecific approaches …, their academic performance improves significantly
And they say this about the data-gathering methodology:
Most of the data currently available on connections between teachers of color and student performance are generated from small-scale qualitative research involving single or multiple case studies. These data focus on a number of significant, though under-recognized, school achievement markers, including attendance records, disciplinary referrals, dropout rates, overall satisfaction with school, self-concepts, cultural competence, and students’ sense of the relevance of school.
While they themselves give the caveats “small-scale” and “qualitative” and don’t seem to consider other, more conventional markers such as grades and rates of conversion to higher education, this looks like it might be a valid conclusion.
However, the report doesn’t provide an itemisation of the cited studies (which, by the way, is inexcusable). A lot of the evidence seems to have been collected in a monograph . Unfortunately I’m unable to check this out myself since I couldn’t obtain access to it online.
-  Gay & al., “The Presence and Performance of Teachers of Color in the Profession”, 2003.