Teachers from the top third? Third of what?

McKinsey & Company's new report, Closing the talent gap: Attracting and retaining top-third graduates to careers in teaching, asserts that the world's top-performing nations "recruit 100% of their teacher corps from the top third of the academic cohort." The support for that claim comes from Exhibit 3 in the report, which shows that:

• Finland has "all teachers recruited from top 20% of high school academic cohort."
• South Korea has "Primary school teachers recruited from top 5 percent of high school academic cohort."
• Singapore has "All teachers recruited from top 30% of high school academic cohort.

Notice the word "high school."  The data on other countries is about high school graduates.

However, McKinsey & Company may be suggesting that the U.S. should try to recruit from the top third of college graduates.  I say "may be" because the report systematically simply says "graduates" in each place that it makes a recommendation.  That could mean high school or college.

One reason to think college graduates are meant is that the report says that in the United States,"only 23% of new teachers overall –and about 14% of those in high-poverty schools– come from the top third of graduates."  That finding is "derived from the US Department of Education, NCES, 2001 Baccalaureate and Beyond Longitudinal Survey," which only included college graduates, In my first attempt to figure out how the report used the survey, I could not find any data that was gathered about high school rank or GPA.  Though I may learn more and have to revise this post, I think the the 23 percent finding is about U.S. teachers as a percent of college graduates.

A second reason is that the report bases an entire chapter on "McKinsey's market research with 900 top-third college students."  That, too, suggests that the report's proposal is that the United States should recruit its teachers from the college top third.

Going beyond the report itself, I did some additional looking at key numbers.

To use only the top third of college students to fill teaching vacancies, the U.S. would need for 24 percent of those students to go into teaching –and that is assuming they would stay for an average of 25 years.  Kentucky, with a lower rate of college graduations, would need 27 percent of top-third college graduates to go into teaching.  By comparison, filling the same jobs from the top third of high school graduates, both the country and the state would need roughly 12 percent of those students to enter teaching.  Drawing from high school's top third seems much easier to accomplish, and since that seems to work for Finland and Singapore, it might work for us, too.

It is also worth asking what puts a student in the top third of high school graduates.  To think about that, I checked ACT results for the six states that test 100 percent of their students and found that for 2010:

• 31 percent of Kentucky high school graduates had an ACT composite of 22 or higher, as did 33 percent of Tennessee graduates,  34 percent of Michigan graduates, and 35 percent of Wyoming graduates.
• 35 percent of Illinois graduates and 36 percent of Colorado graduates had ACT composites of 23 or higher.

So, if Kentucky teacher preparation programs regularly require a 21 composite for admission, it is possible that we are already within shooting distance of the top third of high school graduates becoming our teachers –not there, but in reach with moderate effort.  Naturally, there are other ways to figure out who's in the top third in high school, and it would be great to have our new longitudinal data system give us an idea of how close our teachers already are based on those other measures. 

All of which leads me to end with these core questions:  Is McKinsey & Company recommending that we  aim to draw all teachers from the top third of college graduates?  If so, what evidence does the firm offer to support their recommendation?  

Sources: Digest of Education 2009 statistics provided the numbers for my estimates of needed teachers. The high graduate counts are for 2006-07, combining public school data  from Table 104 and private school data from Table 62.  The college graduate count is 2007-08 data from Table 321, and the 2007-08 teacher counts are from Table 65.  ACT composite percentages come from the 2010 state profile reports.