The "Differentiated Compensation in Education" conference, hosted by fellow Buckeye Staters Battelle for Kids, in Houston this week reminds me how messy the "nuts and bolts" of policy implementation can get. Performance pay to teachers??imposes incredible challenges in the way of program evaluation, validity of test scores, political buy-in, sustainable funding, and unintended consequences. While these are all tough issues to sort through, I am jealous that Texans have the opportunity to grapple with them. Texas has had three major statewide performance-pay programs (see a 2007 report on the first two, GEEG and TEEG, here), Houston's ASPIRE program, and well-known merit plans in districts like Lamesa, Dallas, and Austin that go back as far as 1995.
Yesterday district officials, principals, and teachers participating in District Awards for Teacher Excellence (DATE) (a $397.5 million statewide grant program) outlined the goals and structure of their particular performance pay program. The descriptions were stunningly diverse - there seem to be as many ways to spend DATE money as there are belt buckles in the Houston airport.
Part 1 of the DATE funding (which represents about 60 percent of the overall grant) goes to principals and teachers, but how that is allotted is entirely up to the district. Awards may be available to all teachers, teachers only in tested subjects and grades, teachers at the poorest schools within a district, or subject/grade-level teams. While one district official defends targeting awards to teachers only in tested grades 3-11 (I'd hate to be a second-grade reading teacher there!), a few others seem to understand the perverse incentives this imposes and have designed special ways to give awards teachers who "contribute" to test scores (e.g., K-2 and inclusion teachers).
Part 2 of a district's DATE grant represents "everything else" and can be spent on: data-building, professional development, grant management, mentoring, or even bonuses to cafeteria workers. While diversity in spending and the general drumbeat of experimentation is exciting, program evaluation must be a nightmare. If student test scores improve, how will districts know what to attribute it to? To the performance bonuses, or to byproducts of the awards-such as improved retention, collaboration among teachers, or attrition of the worst teachers? Or to tutoring, professional development or other components having nothing to do with performance pay whatsoever?
For now, judging DATE in its second year is difficult and we can only speculate as to whether it will improve student test scores better than its predecessors. The Institute for Public School Initiatives (a conference co-host) says that by May 2009, 92 percent of participating schools reported an increased focus on collaboration due to the program; 93 percent reported better use of data, and one half of teachers said they used state standards more than before. This tells us that incentives change behavior but it says nothing about whether these changes improve student achievement. How important are these intermediate variables? Should we use them to judge the success of performance pay programs? And what does it say about performance pay as a national reform tool if even Texas has a hard time with these questions?