Not long ago, Florida Gov. Charlie Crist put the last nail in his GOP-Senatorial campaign coffin by vetoing state legislation that would have implemented wide-ranging teacher performance pay. Crist said the bill was “rushed through” and contained “significant flaws;” under the proposed legislation, Florida teachers’ pay would have been tied directly to student performance on standardized test scores.
If Floridians valued a thoughtful approach to spending tax dollars, they might have applauded Crist’s decision from both sides of the political aisle. From the left, tying teacher pay to a single standardized test score has never been popular, as evidenced by the amount of pushback the legislation received from teachers, parents and students. But from the right, fiscal responsibility is supposed to be a foundational tenet of conservatism. When you aren’t sure what you’re buying, it’s a good (and conservative) idea to hold off.
And that’s the issue with merit pay tied to test scores at present. We’re not exactly sure what we’re buying. We think that if we pay teachers more, they will work harder, or smarter. Our extra tax dollars will leverage change. But what if (God forbid!) it works? Where will the continuing increases in tax dollars come from — is the program sustainable? In this economy, can we bet large sums on something we aren’t sure will work, and we are sure that — if it does work — will cost increasing sums year after year? What if, after five years, we have to back off? If the plan assumes that teachers work harder for more money, might the reverse be true — might they stop working so hard if those incentives wither or decrease down the road?
And what if we have the actual metrics of school improvement wrong? While no one can honestly argue against the idea of identifying excellent teachers and then recognizing and rewarding them, how we go about that process needs to be well-thought out and account for the complexities of schooling more than the Florida plan does.
Every major teaching and learning organization, from the American Psychological Association to the National PTA, warns against using a single test score in isolation for any high-stakes decision, and determining which employees get financial bonuses and which do not is certainly high-stakes. Yet the Florida merit-pay model hinged on whether the kids in a teacher’s classroom improved their state test scores in a given year.
Proponents would argue that the test score is one of many factors, but the plan clearly required that merit pay be tied to evaluations, and that evaluations largely (“more than 50%”) be determined by test scores. In a comprehensive, thoughtful plan, achievement as measured by test scores should be only one of many important indicators of teacher quality.
The plan did permit local districts to include other factors in their evaluation processes, but no matter what other factors were selected locally, the Florida plan required that districts submit to the state only the percentages of teachers rated “unsatisfactory,” “needs improvement,” “effective” and “highly effective” and the student performance (test scores) of their students. This is ostensibly so the state can “check” to see if the district got the evaluations right. Without those additional data points included by local districts, the state would have been left to examine teacher evaluations solely in terms of test scores.
Merit pay plans like the Florida plan also have a distribution problem: Who gets the reward for any single child’s success? Should the bonus go to Mike’s math teacher (that’s where he improved the most!), or to his PE coach, who made Mike show her his homework every morning before class, and helped him make corrections until Mike wanted to please the coach as much as he wanted to please himself? Maybe his reading teacher deserves the money since she helped him decode the lengthy word problems on the state test. Or maybe it should go to Micaela, the social worker, who made sure Mike had breakfast and lunch every day, and helped Mike’s parents arrange for the eye exam and glasses he needed. Or, realistically, maybe the bonus should go to the art teacher, who ran a math tutorial session leading up to testing week, and showed Mike how to sketch out diagrams to solve his math problems when he got stuck.
Merit play plans hinge on the responses of individuals to potential rewards, but if the plan spreads out to reward all of these persons — how is that substantively different than using group rewards? The Florida plan as presented didn’t evidence that it could account for these complexities.
Some parts of the Florida plan were sensible. For instance, paying teachers more to work in high-needs schools is a logical investment, because that could result in more expertise and stability for a campus. But these elements weren’t the centerpiece of the Florida plan. In fact, the centerpiece of the Florida plan was loss of local autonomy: The state, rather than local school boards, would have examined and approved district-level student exams, teacher pay plans and the dispersion of teacher evaluation ratings.
By demanding a more thoughtful plan, Charlie Crist effectively said, “Let’s make sure the money we spend will get to the right people for the right reasons, and guarantee some return on investment.” He also rejected the concentration of power to approve and disapprove testing, evaluation and teacher pay at the state level. For these kinds of fiscally and traditionally conservative positions, he was (ironically) abandoned by Florida Republicans this primary season. Whether Independent-minded Floridians pick up the slack for his commonsense leadership in November remains to be seen.