President Obama’s path to performance pay
In his first major education speech as a presidential candidate, Senator Barack Obama affirmed his support of teachers unions. “I believe in collective bargaining, and I believe that any time you’re talking about wages, workers have to be at the table,” he said in a July 2007 speech to the National Education Association (NEA).
Less than two years later, in his first major education address as president, delivered to the Hispanic Chamber of Commerce in March 2009, Obama explicitly backed paying teachers for performance, a reform the unions vehemently oppose. “Too many supporters of my party have resisted the idea of rewarding excellence in teaching with extra pay.”
Behind this seeming contradiction on performance pay is a complex set of political and policy strategies. Obama and his team are caught in the narrow channel between two important Democratic constituencies: establishment organizations that are opposed to performance pay and the increasingly prominent education-reform crowd that generally supports it. And while the administration appreciates the merits of differentiated teacher pay, this is but one of many teacher-quality policies it hopes to change.
The public record reveals how the administration has navigated these shoals, setting a new course for the federal government’s role in the reform of teacher pay. As senator and president, Obama has made known his education-reform commitments and hesitations in speeches to both unions and business groups. The inclinations of his secretary of education, Arne Duncan, are evident from actions he took while serving as head of the Chicago public school system. Finally, the administration’s handling of two prominent federal programs, the Teacher Incentive Fund (TIF) and the Race to the Top (RTT), offers important clues to the strategic thinking at work.
My analysis of this history has led me to two conclusions. First, though the administration’s apparent tentativeness on performance pay can be partially explained by its deference to organized labor, a larger factor is its interest in creating a new and comprehensive framework for advancing teacher quality. Second, the administration’s strategy for generating change through a combination of incentives, collaboration, and optional reforms did not initially bear much fruit for performance pay, but it may reap benefits over the long term, both for performance pay and for other teacher-quality issues.
Developing a Position
In Senator Obama’s 2007 speech to the NEA, he gave an establishment-friendly interpretation of recent education-reform events. He called No Child Left Behind “one of the emptiest slogans in politics” that amounted to “fill[ing] in a few bubbles on a standardized test.” He vigorously supported an active role for unions in education and said that teacher salaries should be raised across the board.
But he also said that schools should be open to paying more to teachers in tough-to-staff subjects, to those who take on additional work, and to those helping students excel academically.
Politically, this equivocation was savvy: he buttressed his liberal bona fides while nodding toward reform. But it also foreshadowed the challenges his administration would face in trying to run the performance-pay gauntlet by staying in the middle of the road.
In the speech, he attempted to reconcile his support for both sides by arguing that differentiated-pay programs should move forward but that they should be created in collaboration with teachers, not imposed on them, and that such programs should never be based on “some arbitrary test score.”
This raised difficult questions: How do you fairly implement a differentiated-pay plan without empirical measures of student performance, and what if organized labor refuses to accept performance pay at all? The first question would eventually be addressed diplomatically by his education secretary; the second lingers on to this day.
One year later, with the election drawing near, Obama again spoke at the annual meeting of the NEA. He was in no position at this time to reveal how the circle was to be squared. In fact, passages specifically related to compensation were either unusually clumsy or cleverly delusive. He said superb teachers should be rewarded through “better pay across the board.” One spectacularly oblique sentence left muddled whether a teacher should be rewarded for learning new professional skills or raising student achievement and whether that reward should be praise or compensation. He was, however, firm that pay systems should be developed with, not imposed on, teachers.
After entering the White House, President Obama felt less need to dissemble on the subject. In his March 2009 speech to the Hispanic Chamber of Commerce, he intimated that his administration would not only support retention bonuses and additional compensation for teaching in hard-to-staff schools and subjects, but also pay increments for those able to measurably influence academic growth. “Good teachers will be rewarded with more money for improved student achievement.”
Left undecided, however, were the role of the standardized “bubble” tests and the implications of union opposition.
Secretary Duncan refined the administration’s position before the NEA in July 2009. Billed as a “challenge” to the union to “think differently” about job security, evaluations, and more, the speech also revealed that the administration was beginning to think holistically about the policies affecting the teaching profession.
Duncan began by acknowledging the wide distribution of teacher effectiveness. Current practices, the secretary argued, unfortunately treat “all teachers lik e interchangeable widgets.”
To gain a better understanding of variations in teacher quality and then make use of this information, we need improved teacher evaluation systems, Duncan argued. Those currently in place are “deeply flawed.”
Then Duncan opened the door to the use of empirical measures of student achievement in teacher evaluations and therefore, presumably, in teacher pay and other personnel decisions. While acknowledging that today’s “tests are far from perfect” and that “the complex, nuanced work of teaching” can’t be fairly measured by “a simple multiple choice exam,” the secretary defended the use of test scores.
Though they “alone should never drive evaluation, compensation, or tenure decisions…to remove student achievement entirely from evaluation is illogical and indefensible.” Duncan was beginning to sketch a new framework for teacher policies, one that integrated student performance data, teacher evaluations, and a range of personnel decisions, including compensation. In time, this shift would prove to be consequential.
But Duncan also echoed his boss’s deference to labor. “The president and I have both said repeatedly that we are not going to impose reform but rather work with teachers, principals, and unions to find what works.” This hedge would also prove consequential.
A good deal can be learned about both the roots and implications of the Obama administration’s evolving position on performance pay from Arne Duncan’s experience with the federal Teacher Incentive Fund (TIF) in Chicago (see sidebar) and the administration’s efforts to fund and reform the program since 2009.
Since 2006, the federal government has funded a small program to support differentiated compensation, the Teacher Incentive Fund. Developed by the Bush administration, TIF provides funding on a competitive basis to states and districts that implement performance-pay programs for teachers and/or principals in high-need schools.
In its first years, TIF had several strikes against it. It was a new program during a period of domestic budget austerity. It sought to advance what was still a politically contentious policy. And it was advocated by an unpopular administration facing a Congress controlled by the opposition party. Accordingly, Congress never fully embraced TIF during the Bush years, and that administration’s annual budget requests (ranging from $100 to $500 million) were never fully funded. Appropriations were generally just under $100 million each year, a relatively small amount for a federal education program.
The Obama administration could have taken the knife to this Bush-era initiative as it has with the school voucher program in Washington, D.C. Instead, it sought to expand TIF by both seeking increased funding and embedding it in a newly proposed, larger program tentatively called the “Teacher and Leader Innovation Fund” (TLIF).
The 2009 federal stimulus package, known as the American Recovery and Reinvestment Act (ARRA), provided TIF with an additional $200 million (on top of its $100 million regular appropriation for that year).
Coming on the heels of the ARRA was the administration’s 2010 budget request, in which the Obama team proposed nearly $487 million for TIF, more than the Bush administration had requested since the program’s inaugural year in 2006. Congress proved receptive, providing $400 million, by far the program’s largest regular annual appropriation.
In its FY2011 request, the first real opportunity for the administration to put its full mark on the federal budget (since the 2010 proposal went to Congress shortly after Obama was sworn into office), the U.S. Department of Education sought to significantly change TIF by including its priorities in the new, broader TLIF program. The $950 million request was approximately double the previous year’s.
According to administration documents, if created, TLIF would support the expansive category of state and district efforts to develop “innovative approaches to human capital systems.” Though differentiated pay would be a core component of the program, TLIF would also support efforts to increase the number of effective teachers, more fairly distribute high-quality teachers among differently resourced schools, improve educator-preparation programs, develop additional professional opportunities for effective teachers, strengthen evaluation systems, remove ineffective teachers from the classroom, improve professional development, and support school turnaround efforts.
So what is to be made of the Obama administration’s initial embrace of TIF and subsequent inclusion of many of its objectives into the TLIF proposal? What does this tell us more broadly about the administration’s views on and intentions for performance pay?
Two different interpretations seem plausible. The first is a political explanation. By supporting TIF, both the Bush-era version and even more so the amended TLIF version, the administration can keep one foot in the reform camp and another in the establishment camp.
Even the original TIF program allows for a wide array of approaches to differentiated pay, some of which opponents find easier to swallow than others, like those that reward all adults in a school rather than just the teachers who measurably increase student performance. TIF also permits grantees to apply program funds to a range of more traditional activities, such as professional development and data collection. This list of less controversial activities would grow under the proposed TLIF initiative. Both programs are optional, so no district or state is required to differentiate pay. Finally, since the program is directed toward high-need districts and schools, most of which have collective bargaining agreements, a state’s or district’s participation in the program ordinarily means that organized labor was involved in crafting the new arrangements. The administration can claim the mantle of reform while standing by its pledge that reform will not be forced on teachers and their unions.
A second interpretation is that the administration is attempting to develop a new, comprehensive federal approach to improving teaching, one that combines student performance data, teacher evaluations, and a host of personnel decisions. The roots of this approach can be seen in Secretary Duncan’s TIF experience in Chicago and in his 2009 NEA speech.
This interpretation is supported by TIF draft regulations released by the education department in early 2010. Among other things, the agency sought to require grantees to measure student growth and use these data in robust teacher evaluations, which would then be aligned with professional development. Language in the administration’s 2011 budget description of the new TLIF implied that TIF was too myopic, treating performance pay as a discrete activity when, instead, policy should reflect the “interconnectedness” of compensation reform and other teacher issues. TLIF, according to the budget document, recognized that it is “important to think of [these issues] in a coherent, integrated way.”
So which interpretation better explains the Obama administration’s approach? Support for both can be found in the administration’s signature program, the Race to the Top.
Race to the Top
Included within the ARRA’s nearly $800 billion in spending was the largest competitive grant program in U.S. Department of Education history, the $4.35 billion Race to the Top (RTT).
The official RTT application was a blend of reform and deference to the establishment. The four major ARRA reform categories—data use; standards and assessments; failing schools; and teacher quality—served as its backbone. But the administration added a good bit of muscle. States would earn points for having in place each of the 12 data elements required by the federal America COMPETES Act. They’d be rewarded for having policies authorizing aggressive interventions for failing schools. They’d be significantly penalized for lacking a charter school law. And they’d be barred from even applying if they had “data firewalls” preventing student performance information from being tied to individual teachers.
But states also earned significant points for crafting plans that earned the blessing of their school districts and unions. In a number of cases, those who scored state applications gave extra weight to stakeholder “buy-in” by subtracting points from proposals that lacked the support of these groups.
Though Duncan would later downplay the importance of consensus, when Delaware and Tennessee were announced as the only first-round winners the secretary emphasized that these two states stood apart in their ability to develop strong proposals that also had broad support. In fact, the most hotly debated RTT question in the spring of 2010 was how states would address the tension between reform and union buy-in in their second-round applications.
At first glance it is striking, even startling, how small a role performance pay played in round one of Race to the Top. The application has six main sections: one for each of the four ARRA reforms; an introductory section largely dedicated to buy-in issues and previous reform successes; and a final catchall section.
The fourth section (D), “Great Teachers and Leaders,” contains the most points of the six (138 out of 500, or 28 percent). It is broken into five subsections, one of which is titled “Improving teacher and principal effectiveness based on performance.” This comprises four sub-subsections, including “Using evaluations to inform key decisions.” That is broken into four sub-sub-subsections, one of which includes performance pay. Performance pay is one of three elements in this area, along with promotion and retention.
In other words, in the Race to the Top, performance pay is a sub-sub-sub-subsection.
Were a peer reviewer to score by the book, a state without a performance-pay plan would lose just over 2 points out of 500. By comparison, a state without a charter law would lose 32 points.
The most straightforward interpretation is that the administration capitulated to performance-pay opponents. But this analysis seems incomplete, even unfair. Had pleasing the establishment been the administration’s priority, it might simply have kept performance pay out of the application altogether.
In fact, subsection (D)(2) offers compelling evidence for the alternative interpretation. It asks states to measure student growth and to tie these results to individual teachers. It also asks states to develop annual teacher evaluations and include student growth as a component of each teacher’s official assessment. Finally, it asks them to use these evaluations to inform a number of personnel decisions, such as tenure, removal, and compensation.
The Obama administration appears to be offering a new—not to mention tight and rational—framework for improving the teaching profession. However, consistent with the administration’s nonconfrontational method for advancing reform, the new framework is optional. Since RTT is a competitive grant program, no state is forced to participate; states uncomfortable with the framework are free to disregard it.
It is too soon to tell whether this new framework will lead to better student outcomes. But it is not too soon to test the administration’s theory of action for bringing about change. Did the Race to the Top’s use of financial incentives, rewards for collaboration, and optional reforms lead to progress in performance pay and other policies that affect the teaching profession?
State Race to the Top Applications
In the first round, 40 states and the District of Columbia submitted Race to the Top applications. To test the effectiveness of the Obama administration’s approach, I reviewed each application’s (D)(2) section. Figure 1 illustrates how many proposals include an affirmative response to the nine questions embedded in the RTT framework. Will the state…
1. measure student academic growth?
2. conduct annual teacher evaluations?
3. include student growth in teacher evaluations?
4. use teacher evaluations to inform professional development decisions?
5. use teacher evaluations to offer additional professional opportunities?
6. use teacher evaluations to inform compensation decisions (performance pay)?
7. use teacher evaluations to inform tenure decisions?
8. use teacher evaluations when considering promotions?
9. use teacher evaluations to inform termination decisions?
Of the 41 entrants, 39 have systems in place to measure student growth, are building such systems, or have committed to building them. Most states (32) also agreed to conduct annual teacher evaluations. In some cases, this represents a major shift in policy; for example, under current practices, tenured teachers in Hawaii are evaluated only once every five years.
Only about half of the states (21) agreed to include measures of student growth in teacher evaluations. Several committed to having 50 percent or more of each teacher’s evaluation composed of such data. A number of states, however, simply ignored this matter in their applications or only committed to forming a stakeholder committee to discuss it.
Almost all states (34) committed to using evaluations to determine which teachers need which types of professional development. But states were far less likely to commit to using evaluations to make tougher personnel decisions. Only nine were willing to link teacher evaluations to processes for terminating the lowest-performing teachers.
Sixteen states committed to performance-pay plans. But only five states proposed what could be considered strong plans (Arizona, Delaware, Florida, South Carolina, and Washington, D.C.). Notably, Florida required all LEAs (local educational agencies) participating in the state’s application to make student achievement growth the most significant component of compensation, ahead of years of experience and academic degrees.
Two plans could be considered of moderate strength. Minnesota planned to expand its “Q Comp” program, but nearly all details were to be negotiated at the local level between unions and districts, raising questions about the ultimate impact of the plan. In Georgia, participating districts agreed to adopt ill-defined step increases for high-performing teachers.
The remaining nine performance-pay plans were of dubious seriousness. In several applications, including Oklahoma’s and West Virginia’s, the state promised to create a bonus pool but made district participation optional, so it is possible that no teacher would receive extra pay based on merit. In Massachusetts, 1 percent of the state’s districts would pilot a locally determined, yet-to be-defined, differentiated compensation plan. In Idaho, all employees of schools in the top three quartiles of statewide student growth would receive small bonuses, meaning half of the state’s below-average schools would get schoolwide bonuses.
Before choosing Delaware and Tennessee as the first round winners, the education department identified 16 finalists. Figure 2 shows how committed the group was to key components of the new teacher-effectiveness framework.
So far, RTT has not had a revolutionary impact on performance pay. This seems to raise questions about the administration’s belief that large federal financial incentives will lead states to embrace controversial reforms. The limited use of student-performance data in teacher evaluations offers further evidence for this point. Although states, in order to apply, had to remove data firewalls, only half of applying states took the critically important but optional next step: actually making student growth a part of evaluations.
An additional data point calls into question another component of the administration’s theory of action—that major reform can be brought about through collaboration with unions. As noted above, five applicants proposed strong performance-pay plans. South Carolina has no teachers unions. Washington, D.C.’s proposal received no union support. In Florida and Arizona, 8 and 21 percent of local teachers unions, respectively, supported the state’s plan. Only Delaware was able to both craft a strong performance-pay plan and earn broad union support (100 percent).
A Solid Footing
Several factors have diluted the administration’s work on performance pay. First, Duncan, as a general rule, prefers to make reform optional, using incentives to alter behavior. Second, the secretary appears to be more interested in changing the teaching profession broadly than in advancing the narrower issue of performance pay. Third, and most important, the president and secretary remain committed to securing union support for change, reform “with” labor not “to” labor. RTT winner Tennessee made alternative compensation systems completely optional for districts and required that, before a local performance-pay plan is implemented, it receive the blessing of the local union.
But it may still be the case that, in the long term, the administration’s efforts will have a profound positive impact on performance pay. A few states were willing to consider performance pay to an extent that they hadn’t before. And while giving unions a great deal of power in negotiations about differentiated pay will severely limit the number and strength of plans adopted, it might help ensure the strength and sustainability of the few plans that do emerge. Finally, by encouraging states to measure student growth, embed student learning in annual teacher evaluations, and use evaluations to inform a range of personnel decisions, the administration has laid the foundation for performance-pay plans in the future.
The Obama administration, if nothing else, has changed the politics of performance pay. No longer can it be assumed that leading Democrats will oppose efforts to financially compensate high-performing teachers.
On January 19, 2013, in other words, we’ll be able to ask of a Democratic administration a once inconceivable set of questions. How many billions did it spend on performance pay? How many new state-level performance-pay plans did it bring about? Did its activities cause unions to drop their reflexive opposition? Is performance pay now widely viewed as one part of an integrated teacher policy framework?
Depending on the answers to these questions, performance pay and the new teacher framework—not turnarounds, a reauthorized ESEA, or another higher profile issue—may be the Obama administration’s most important education legacy.
Andy Smarick, a former U.S. deputy assistant secretary of education, is an adjunct fellow at the American Enterprise Institute and the Thomas B. Fordham Institute.