What's a teacher worth?

The "teacher famine," complained New York University education professor Adolphe Meyer, was caused by "economic cheeseparing.... First-rate men and women have long since turned their back on chalk and blackboard; but now it is becoming difficult to attract even those of a lesser caliber." If the complaint sounds dated, it is. Meyer was writing in 1957. Plenty of things have changed since then, including the total amount of money we spend on public education--now, almost $11,000 per student, more than any other country in the world. "Cheeseparing" today comes less from a parsimonious public than from an inefficient and ineffective salary distribution system, as our three Forum contributors--Lewis Solmon, Brad Jupp, and Julia Koppich--suggest. It is not that average salaries are terrible, but that our best teachers are paid little more than our worst. What is remarkable is that Solmon, a former education dean, Jupp, a union leader, and Koppich, a "new union" advocate, agree that the debate is no longer whether to throw out the single salary schedule by which most of our teachers are paid, but what to replace it with. On the pages that follow, they offer provocative--and concrete--proposals for paying teachers what they are worth while providing students with an education they deserve. As Jupp says, "We are in an exceptional moment, one where the single salary schedule can no longer support the pressures placed on it."

[ILLUSTRATION OMITTED]

[ILLUSTRATION OMITTED]

The Uniform SALARY SCHEDULE

A Progressive Leader Proposes Differential Pay

For at least two and a half decades, political leaders and opinion makers have been telling teachers and union leaders like me that it is high time to move away from the single salary schedule. For a long time it was easy for us to dismiss those calls for change. This was partly because as a profession we are more remote from the policy debate than we should be, but it is also because we believed that many of those engaging in the debate were ax grinders, more interested in dismantling public education as an enterprise than improving it.

Recently, though, with more voices joining the choir calling for change, it has become difficult to write off the differentiated pay advocates. The 1996 report of the National Commission on Teaching and America's Future, "What Matters Most," was a wake-up call. Here was a report that placed the highest value on the teacher, even as it recommended "develop[ing] a career continuum and compensation systems that reward knowledge and skill." Since then, other important pro-teacher groups, like the Progressive Policy Institute, Public Agenda, and the Teaching Commission, have taken even stronger positions. In his paper "Better Pay for Better Teaching," Bryan Hassel, who is a researcher for the Progressive Policy Institute, raises the stakes by saying, "In addition to better pay, we must move toward a better pay system. We should reward teachers not just for experience, but for their skills, knowledge and, ultimately, the performance they bring into their classrooms."

Less obvious was the pressure from within the ranks of our profession. With the exception of Public Agenda's report, "Stand by Me," I have found little research on teachers' perceptions of alternative compensation. In Denver, however, my union, the Denver Classroom Teachers Association, which represents approximately 4,500 teachers, and independent researchers commissioned jointly by the union and the Denver Public Schools, developed an extensive body of perceptual data that showed teachers as open-minded to changes in the ways they are paid.

Surprising Data from Teachers

Though Denver had a typical salary schedule (see Figure 1) our data overthrow many of the preconceived notions held by teacher unions, school administrators, policy leaders, and opinion makers about how teachers perceive compensation systems. Since 1998 our union has asked its members what they thought about incentives for "teaching at schools with the highest percentage of high-need students." By 2003, when the last available survey was conducted, the number of people favoring these incentives had reached 89 percent. The percentage of teachers who favor incentives for "teaching in content areas of short supply" is only slightly less, at 82 percent.

The union has also asked its members, "What percentage of your pay should be based on accurately measured student growth data?" Respondents were given multiple-choice options of none, 1 to 5 percent, 5 to 10 percent, 10 to 25 percent, and 25 to 50 percent. In 2002 and 2003 no single choice received more than half of the responses, but the fact that fewer than half of the teachers surveyed selected the first choice, none, is remarkable; it means that for two years' running more than half of the union members surveyed believe that some portion of their pay should be based on accurately measured student growth.

In fact, Denver teachers have shown surprising open-mindedness about merit pay programs. The Denver Public Schools, with the collaboration of the teacher union, launched a Pay for Performance pilot program in 1999 and, when it ended in 2003, started a more comprehensive Professional Compensation System for Teachers (ProComp). Our independent researchers discovered a surprising amount of support for merit pay by teachers in both programs.

The hallmark of the Pay for Performance pilot was paying teachers $1,500 bonuses for meeting measurable objectives set collaboratively with their principals and based on the academic growth of the students they taught. When asked in the spring of 2003, just as the pilot program was ending, to rate whether setting measurable objectives for bonuses of up to $1,500 had an impact on "cooperation among teachers," 53.2 percent of the participating teachers said the impact was positive; only 2 percent said the impact was negative. While this positive response is certainly dependent on the special nature of the objective-setting process in Denver--a process in which teachers collaborated directly with their principals to set goals based on individually measured baselines for the students they taught, in the subject matter they taught--this response still flies in the face of preconceptions that teachers fear pay for performance based on student growth because it will harm collegial relations.

Even more surprising were the results that came from phone polls conducted in January and March of 2003 asking teachers their opinions on ProComp just before the vote to adopt the new program. When we ran demographic cross-tabulations of the results, we found that the teachers most likely to ratify ProComp were male secondary teachers with 13 or more years of service in Denver Public Schools. Though perhaps influenced by the special circumstances ProComp created--it lifted a cap on annual salary increases that, due to our single salary schedule, became effective following the 13th year of service--the results refuted the stereotype of the change-averse senior teacher.

Puttering toward Reform

The common perception that current salary scales are set in stone also overstates the case--at least with respect to the Denver metro area. The salary scales in each of the nearby districts we examined displayed evidence of gradual tinkering in response to political pressure. I have not found rigorous national research on this subject, but comparisons of metro Denver salary schedules conducted during the development of ProComp found, according to independent researcher Douglas Rose, that all the public school districts routinely "showed strong hints of incremental decisionmaking." When experienced teachers, for instance, want additional salary increases, steps are added at the top. And boards of education want to compete in the marketplace for entry-level teachers, so early sections of the salary schedule are artificially inflated. Lower rates of pay are also established for probationary teachers or teachers with emergency licenses. Some districts in our area have even made nominal efforts at reform. Denver has market incentives for teachers of English Language Acquisition. Another district, Jefferson County, experimented for a couple of years with a program granting rewards to teachers for completing approved projects.

For all of the changes that have been slapped onto the existing schedules, however, it is safe to say they amount to nothing more than accommodations to internal and external pressures, small alterations that accomplish little or nothing in the work lives of teachers or the learning of students. These are not the radical new systems called for by the National Commission for Teaching, the Teaching Commission, Public Agenda, or the Progressive Policy Institute.

What the Mile-High City Did

In Denver, we developed ProComp to take our teachers further down the road toward a new form of thinking about compensation (see Figure 2). Compared to the system that it replaces, ProComp is a comprehensive approach to merit pay:

* It ties teachers' pay directly to professional accomplishments, including demonstrated knowledge and skills, as well as to student academic growth;

* Salary incentives are also provided for work in hard-to-staff assignments and hard-to-serve schools;

* A "menu" of extra earning opportunities, available to the entire work force, is spelled out, as are bonuses available to specific job categories;

* Teachers are eligible for every incentive earned--there is no cap on such earnings. A special-education teacher, for instance, teaching in an English Language Acquisition assignment and working in a school with the highest level of students receiving free and reduced-price lunch, would receive three bonuses, one for each of three eligible areas.

We also designed the system so it could be sustained across education careers. Using a financial analyst, we developed a 50-year model so that teachers could count on career earning expectations and not just pick up an extra bonus or two now and then. At the same time we built a system around a commitment to change. ProComp is grounded in a nine-year collective bargaining agreement, instead of the normal one- to three-year cycle. This should ensure that it is not a policy du jour, but lasts long enough to establish and prove itself. And we tied the system to larger district goals in order to provide clear educational incentives while keeping pressure on the bureaucracy to both perform and improve.

ProComp has its limits--it is rooted in compromise; the timeline for implementation is slowed by the need to build district financial and instructional capacity; it does not raise entry pay and provides relatively slow growth in annual earnings; many incentives are too small to really meet demands of the market--but I believe it points in some important new directions.

For people impatient to see the single salary schedule get out of the way, ProComp is not ambitious enough. In the final analysis, I think they are correct. ProComp is not an educational "silver bullet" or even a comprehensive solution to the unsolved problem of how to build a new form of teachers' pay.

We recognize, however, that we are in an exceptional moment, one where the single salary schedule can no longer support the pressures placed on it by the expectations of a 21st-century public education system. Fortunately, Denver teachers recognized those new realities. On March 19, 2004, the proposal was ratified by a decisive margin (59 percent to 41 percent) of union members, with 2,700 of the union's 3,200 voters--well over half of the district's 4,500 teachers--casting votes.

We look forward now to testing what we have developed. We want to use the momentum for change to improve both the teaching profession and the schools where our teachers work.

Denver's Salary Schedule (Figure 1)

With relatively low starting salaries and guaranteed raises over time,
the current Denver Public Schools salary schedule is typical of
compensation schemes for teachers. Each step represents a year of
teaching.

          B.A.     M.A.     DOCTORATE

New Hire  $31,320  $31,779
Step 1    $32,971  $33,454  $39,169
Step 2    $33,073  $33,697  $40,903
Step 3    $33,225  $35,101  $42,642
Step 4    $33,480  $36,503  $44,377
Step 5    $33,785  $38,053  $46,251
Step 6    $33,988  $39,671  $48,219
Step 7    $35,421  $41,337  $50,290
Step 8    $36,912  $43,087  $52,449
Step 9    $38,456  $44,924  $54,702
Step 10   $40,092  $46,860  $57,057
Step 11   $41,784  $48,843  $59,521
Step 12   $43,566  $50,944  $62,082
Step 13   $45,546  $53,401  $64,919

SOURCE: Denver Public Schools

Merit and Battle Pay in Denver (Figure 2)

Denver teachers hired before 2006 have a choice between the traditional
salary schedule and this four-dimensional merit pay system. Teachers
hired after January 1, 2006, will automatically enter the new system.

Learning Gains                       Evaluation
* Teachers who exceed expectations   * Teachers found to be
  for student growth as measured by    unsatisfactory will have their
  a statewide Colorado test will       salary increase delayed for a
  receive a sustainable 3% raise.      minimum of one year.
* All teachers will set two student  * Probationary teachers will be
  growth objectives with the help      evaluated every year in their
  of their supervisors. Teachers       first three years of service and
  who meet both objectives will        will receive a 1% raise if they
  receive a 1% raise; those who        are judged to be satisfactory.
  meet one objective receive a 1%    * Non-probationary teachers will be
  bonus.                               evaluated every three years, and
* Teachers at schools identified as    will receive a raise of 3% if
  distinguished will receive a 2%      they are deemed satisfactory.
  bonus.

Battle Pay                           Credentials
* Teachers working in assignments    * Teachers with active licenses
  identified as hard-to-staff and      from the National Board for
  in schools termed hard-to-serve      Professional Teaching Standards
  will receive a 3% bonus.             (NBPTS) will be awarded a salary
                                       increase of 9%.
                                     * Teachers who complete one
                                       Professional Development Unit
                                       (PDU) in their concentration will
                                       receive a 2% raise.
                                     * Teachers who complete an advanced
                                       degree relevant to their
                                       assignment will receive a 9%
                                       raise.

SOURCES: Denver Public Schools and Denver Classroom Teachers Association

BY BRAD JUPP

Brad Jupp is a member of the Denver Classroom Teachers Association and the Teacher/Coordinator of the ProComp Project.

All Teachers ARE NOT THE SAME

A Multiple Approach to Teacher Compensation

It is by now a familiar story, often told as a lament: teachers in this country continue to be paid according to the single salary schedule.

They accrue better pay on the basis of years of experience and college units earned. Units may or may not be related to teaching assignment. Some districts have modestly tweaked this arrangement by paying a premium to teachers who earn certification through the National Board for Professional Teaching Standards. But by and large, in thousands of school districts across the United States, the unvarnished standard single salary schedule prevails.

Teacher unions, among the staunchest defenders of the standard compensation arrangement, are often credited--or blamed--with inventing this salary calculus. In fact, the classic teachers' salary arrangement is an artifact of civil service. Developed in the early 1920s, the system was popularized three decades later as a way of creating salary equity between elementary teachers, most of whom were women, and secondary teachers, most of whom were men. This was not a pre-feminist revolution so much as a necessary economic response to the post-World War II enrollment boom. Over time, to be sure, teacher unions have come to defend the standard single salary schedule in the name of employee equity and fairness.

The National Education Association assiduously avoids anything that might be construed as "merit pay." The position taken by the American Federation of Teachers is less rigid, yet replete with caveats to protect against anticipated slights and abuses. The unions' position is not unwarranted. Merit pay schemes that have been tried in education have an abysmal track record (see "Dollars and Sense," p. 60, for one that had a decent record--while it lasted).

Part of the problem, however, has been that merit pay systems are rarely based on objective standards, a flaw that often created unhealthy competition for the usually scarce resources rather than cooperation among teachers. These insufficient funds all too typically forced many teachers to take their rewards solely in the form of psychic remuneration.

Compounding the compensation dilemma, policies promoted by both districts and unions have endeavored to maintain the fiction that "a teacher is a teacher is a teacher." Compensation structures have failed to recognize that some teaching jobs are more difficult than others or that some teachers are more--or less--skilled than others.

Salaries for the Real World

The time has come for school districts and teacher unions to take a different tack. It is time to develop and implement a professional compensation arrangement that recognizes the complex nature of the work of teaching and that compensates teachers for both the difficulty of the assignment and the professional accomplishment that is part of it.

We need a compensation structure that utilizes multiple approaches. These should include paying teachers more for: 1) attaining knowledge and skills that demonstrably contribute to improving student learning; 2) mentoring newer and less skilled teachers; 3) teaching in hard-to-staff schools and choosing difficult-to-staff subjects; 4) producing higher test scores, using a value-added approach.

These ideas fly in the face of long-established tradition. But it's time to reexamine that tradition. It's time to acknowledge publicly that some teaching jobs are more difficult than others. And we must be willing to pay more for some fields than for others. In a perfect world, perhaps, a physics teacher is no more valuable than an English teacher. But we do not live in a perfect world. We live in a world in which physics teachers are at a premium, and for the foreseeable future supply and demand will need to prevail. The market must have its way.

Some will argue that what is suggested above is too complicated, that it is time to scrap the old salary schedule and pay teachers on the basis of their students' test scores alone. After all, assert proponents of this argument, isn't the true measure of a teacher's worth her students' test results? In fact, at the base of this argument lies the same fault line that threatens No Child Left Behind. Making judgments about student learning by simply examining test scores from one year to the next is hazardous at best. Tests provide a simple snapshot in time and may not be well aligned with standards or curriculum. Moreover, particularly in urban districts, given the rate of student transience, the cohort of students tested at the beginning of the year may be different from that tested at the end of the year, thus providing few useful comparative data.

The New Math for Merit

But there is a way to use test scores to gain needed information about the impact of teaching and the levels of student learning: value-added calculations. The value-added approach has the advantage of separating student effects (ethnicity, family background, socioeconomic status) from school effects (teachers, administrators, programs) since it examines test scores to determine if students are making anticipated academic gains each year. Measured on the basis of their progress from the previous year, students, in a sense, act as their own statistical control. Value-added programs calculate a projected test score for a student in a given grade or subject based on his or her previous academic achievement. The difference between the actual score and the projected score is the value added.

Value-added calculations, however, should not be used as the sole gauge of teachers' compensation. They too are an imperfect technology. But they can, and should, serve as one important measure. Consistent value-added work by William Sanders in Tennessee has shown that several consecutive years of teachers' adding measurable value to students' learning provide a foundation on which students can continue to make academic progress. After several years of ineffective teaching, students may never recover academically.

Finally, it is not possible to discuss teachers' compensation without taking up the issue of their evaluation. In most places, evaluation is done poorly, with checklists about behavior standing in for standards of good practice that should frame evaluation systems. Administrators typically in charge of the process have too little time or training to effectively help teachers improve their practice.

But there is an effective alternative to the ineffective evaluation as well. Systems of peer review, in place for a decade or more in a dozen or so districts--such as Toledo and Columbus, Ohio; Rochester, New York; and Montgomery County, Maryland--have shown remarkable promise. Beginning teachers are provided the support they need from specially selected experienced teachers, who are chosen jointly by the district and the union. Those individuals who were not meant to be teachers are soon out of the profession. Struggling tenured teachers are given the support they have long needed. Should that not prove adequate, they are encouraged to find other lines of work. Conventional wisdom notwithstanding, teachers, including unionized teachers, are able to judge their colleagues fairly but rigorously. Yes, some teachers are dismissed. More important, evaluation accomplishes the purpose for which it is intended: improving professional practice.

Working with Complexity

Marrying a well-developed system of peer review and value-added test scores could create a powerful framework for teachers' compensation. Adding pay for knowledge and skills, compensation for mentoring, and pay for teaching in hard-to-staff schools and subjects will transform a pro-forma salary schedule into a professional compensation arrangement that better recognizes the complexity of teaching and offers teachers the kinds of incentives and options that professionals deserve.

Salaries by themselves, no matter how high or competitive, will not encourage teachers to remain at schools where the working conditions are poor. Competent, supportive administrators, a decent physical plant, and requisite instructional supplies are the sine qua non for maintaining a quality teaching staff, regardless of the rate of pay.

In sum, it is time to construct a salary schedule that gives teachers choices, opportunities, and options--pay for knowledge and skills, pay for mentoring, added pay for hard-to-staff schools and subjects, and added compensation for test scores calculated using a value-added approach.

The hope is that progressive unions and districts will take up the challenge to shape this new salary construct. They will come to see rethinking compensation as part of their obligation to promote quality teaching, and as the next step on the road to creating a true profession.

BY JULIA E. KOPPICH

Julia E. Koppich is an education consultant based in San Francisco. A former high school teacher and faculty member in the school of education at the University of California, Berkeley, she is also coauthor of United Mind Workers: Unions and Teaching in the Knowledge Society.

Recognizing DIFFERENCES

Let's Reward the Good Teachers

For more than a century, public education has worked under a single salary schedule that compensates teachers for college credits, education degrees, and years of experience, but not for their effectiveness in the classroom. (See Figure 1.)

In fact, research shows that the degrees, courses, and experience that teachers have, beyond the first few years of teaching, are unrelated to how much their students achieve. Furthermore, the current salary schedule does not normally take into account the fact that teachers work in schools offering different levels of nonmonetary benefits, such as a safe, pleasing environment. Nor does it recognize that students come to class with different levels of preparation and home support.

Paying all teachers with the same experience and credits the same salary also ignores the fact that graduates of different fields have vastly different alternative career options; think of a physicist compared with someone having a bachelor's degree in elementary education. School administrators report that it is very difficult or impossible to fill elementary teaching positions about 6 percent of the time, while positions in math, physical sciences, and special education are difficult or impossible to fill more than 30 percent of the time. According to the American Federation of Teachers, the estimated starting salary for teachers with a B.A., in 2003-04, regardless of discipline, was $30,496. The National Association of Colleges and Employers tells us that the average starting salary for accountants that year was $41,110, and for graduates in computer science, $49,691. (See Figure 2.) Starting salaries for mathematicians with a B.A., the year before, averaged $40,512; for physicists with a B.A. in 2002, according to the American Institute of Physics, $78,000. Is it any wonder that it is so difficult to hire and retain math and science teachers?

Extra pay for those in hard-to-staff fields would acknowledge the laws of supply and demand--greater opportunities for math or science majors bid up their earnings in other careers, so getting them into teaching requires competitive salaries. It would not mean that physics teachers are more important than elementary teachers.

In the current system, an increase for one teacher requires increases for all. If, for instance, we decided that our teachers are underpaid, as state officials from all parts of the country tell me, current practice would make a meaningful pay increase prohibitive. Just to bring the salaries in the below-average states to the national average would cost $8.5 billion--an amount that is fiscally irrational. It may seem like a meaningless argument, except that such an amount is dictated by the current uniform salary schedule, which requires those below-average states to raise each teacher's salary. It would be much cheaper--not to mention more educationally effective--to raise the salaries of just the most-effective teachers, those deserving of the increase, rather than all teachers. Then there would be money available to give larger raises to the very best teachers.

It would be similarly impossible to bring all teachers' salaries up to the average level of other professions. Why would we want to pay more to the least-effective teachers anyway? What we end up with, then, are paltry average annual increases (as teachers gain experience and course credits), ranging from the high of a $1,498 average increase in California to a meager $503 at the low end in South Dakota.

In the Madison School District in Arizona, the lowest salary for a new teacher with only a bachelor's degree is $31,304 and the highest salary after many years and 72 postbaccalaureate credits is $57,251, an 82 percent increase over a career! Compare that with the legal profession, where the lowest-paid 10 percent earn less than $44,490 and the highest 10 percent earned more than $145,600--a 227 percent difference! The flat salary schedule for teachers is a good reason for those in Madison to welcome the stipends of $6,250 that the district offers to its "master" teachers, who take a leadership role among the faculty, and the bonuses averaging $3,400 to teachers who exhibit outstanding classroom performance and student achievement.

The traditional K-12 compensation system is obsolete in that it is no longer useful, but, sadly, it is not obsolete, because it is still in use.

The Teacher Advancement Program

For the past four years, the Milken Family Foundation, founded in 1982 and based in Santa Monica, California, has been working through its Teacher Advancement Program (TAP) in Madison and in more than 60 other schools--mostly elementary and middle schools and a few high schools--around the nation to change the way teachers are evaluated, helped to grow professionally, and compensated.

The TAP is a systemic reform of public schools intended to attract, motivate, develop, and retain high-quality talent in the teaching profession. It has four key elements:

* Multiple career paths allow teachers to pursue a variety of positions throughout their careers--career, mentor, and master teacher--depending on their interests, abilities, and accomplishments. As they move up the ranks, their qualifications, roles, and responsibilities increase--and so does their compensation. When teachers take on more responsibilities, they should receive more pay. The old career ladder programs failed because the best teachers were honored with new titles and more work, but with meager, if any, extra pay.

* Performance-based accountability evaluates teachers' effectiveness through a comprehensive, research-based system that combines such criteria as position responsibilities, classroom observations, and students' gains in test scores.

* Ongoing applied professional growth requires a change in the school schedule that allows time during the regular school day for teachers to learn, plan, mentor, and share with other teachers so they can constantly improve the quality of their instruction.

* Market-driven compensation allows schools to compensate teachers on the basis of their performance and the performance of their students.

Attracting--and Keeping--Effective Teachers

One of the things we all seem to know is that there are few careers, except teaching, in which professionals are not held accountable for their failures and rewarded for their accomplishments. So why, political constraints aside, do we insist on giving raises to reward a teacher's longevity instead of his or her job performance?

For the past 17 years the Milken Foundation has worked with state superintendents to recognize and reward K-12 educators with a $25,000 gift--no strings attached--for outstanding and effective performance. Of the 1,977 recipients of the Milken award, we know that 1,653 of them are still working. Of these, 281, or 17 percent, have left their school buildings to take jobs at a district office (154), a nonprofit organization (47), a university (41), a federal or state government office (24), or a private company (15).

Unfortunately, we do not know whether this 17 percent turnover among Milken award recipients is better or worse than among those equally experienced and outstanding teachers who do not receive the awards. But we do know that 46 percent of all teachers leave the profession in the first five years and that, anecdotally, the Milken awards have created an aura of excellence about the profession that enhances the environment that our TAP research shows is so important to teacher retention. The very fact that these teachers have opportunities beyond the classroom increases the attractiveness of the profession as a whole.

However, when no compensating salary is awarded for teaching in difficult or undesirable schools, it is easy to understand why the best teachers may choose to teach in the most rewarding and pleasant environment available, moving from low to high socioeconomic status schools when the opportunity arises. Extra compensation is thus needed for those teaching in hard-to-staff schools where conditions are difficult, dangerous, or unpleasant. Defying that pattern, however, some very talented teachers in Arizona are moving from socioeconomically advantaged schools that are not using the TAP to schools of low socioeconomic status that are. Over the past three years, 61 teachers have started working at the two schools of lowest socioeconomic status in the Madison school district, both of which are using the TAP. Of these teachers, 13 (21 percent) have come from schools in high socioeconomic areas in Madison or nearby districts, and they are among the best teachers from the area. They are attracted by the more interesting professional development, the enhanced collegiality, and the opportunity to earn more by being effective.

We want teachers who love kids and want to help them learn, but that does not mean they cannot be interested in compensation as well. Physicians seek to prevent or cure diseases, and some lawyers seek to dispense justice, but that does not bar the best of them--but not all of them--from earning large incomes as well.

Student Outcomes Count

Teachers should be rewarded for producing useful student outcomes, most notably, student learning gains, measured by value-added standards (i.e., improvement) rather than by levels of achievement at the end of a course. This method takes into account differences in where students start as well as differences in out-of-school factors that teachers cannot control. Looking at gains rather than levels of achievement also adjusts for the fear that performance pay will make all teachers want to teach the highest-achieving kids. When student improvement is rewarded, there may be financial benefits to teaching students who have the longest way to go--it may be easier to get a 25 percentile gain from someone starting at the 30th percentile than a 15 percentile gain at the 80th percentile. Providing incentives for teachers to make their students learn more may encourage teachers to do so, but, perhaps more importantly, it will compensate them for the extra effort required to improve the skills that will help their students achieve.

Another issue involves incorporating a school's nonacademic goals into a merit pay system. As we know, schools are expected to develop students' social behavior, career preparation, and positive attitudes. The most hackneyed of these has been enhanced student self-esteem, which sometimes is used as a reason for not failing students who deserve to fail. Let it simply be said that a great way to enhance a student's self-esteem is to have her achieve something academically. Holding teachers accountable for students' academic achievement gains is not inconsistent with students' accomplishing other things in school.

Standards without a Straitjacket

Despite the need to keep the focus on academic achievement, the Teacher Advancement Program acknowledges that research has identified pedagogical methods that help students learn, so it includes evaluation of classroom skills as part of its teacher compensation system. This allows teachers to be rewarded if they do everything right, even if their students' scores do not increase.

A key to this part of the performance evaluation, of course, is developing clear criteria for measuring those classroom skills. Multiple evaluations are conducted by certified evaluators. This helps overcome teachers' fears of bias and nepotism in their evaluations. Moreover, by including principals in the school-wide performance bonus system, they too will have an incentive to ensure that the most-effective teachers are rewarded.

Will such rubrics reduce teachers' opportunities to be creative in their teaching, to try new approaches, to teach as they like? Probably not, because the other test of teacher competence is better student outcomes. Thus the rubric gives credit for the skillful exercise of proven teaching methods, regardless of student outcomes, as well as rewards for student success, regardless of teaching method.

Fix the Whole System

The initial success of these performance- and responsibility-based compensation systems suggests that there are alternatives to the traditional "step-and-column" pay in which no one will earn less than in the traditional system. All teachers who reach certain goals get a bonus; but 50 percent of that bonus is awarded for teaching skills (a classroom-based evaluation) that are not tied to student outcomes and 50 percent for student achievement gains that are not part of the teaching skills evaluation. Furthermore, half of the student outcome bonus is based on school-wide gains and half on gains by an individual teacher's students.

In the three years that we have been using the TAP system, we have found that the school-wide rewards part of the bonus encourages teachers to work together to make everyone more effective. There is the possibility of the "free rider," of course: ineffective teachers reaping benefits from the achievements of their more-effective colleagues. Thus we expect part of the bonus to be based on individual teacher results. All bonuses must be significant or the extra work involved in implementing such a system will not be deemed worthwhile. That means, for instance, that a master teacher in Madison, Arizona, could get a bonus worth as much as 17 percent of his or her salary compared with the 2-3 percent bonus that current salary structures usually set as a cap for such expert teachers.

We have learned that performance pay alone is not enough. It must be supported by a strong, transparent, and fair evaluation system, and by a professional development plan that helps teachers to deal with revealed deficiencies and to improve. Teachers may resist evaluation not because they are unwilling to be held accountable, but instead because they fear they do not know what to do to improve student achievement. If professional growth opportunities are available to help teachers improve, the resistance to being evaluated fades.

A professional development program should use student data to identify areas where teachers need help. It should then have teachers help other teachers improve their teaching and student learning. This fosters collaboration and reduces the competitive attitudes that some fear are engendered by performance pay. Indeed, as long as my receiving a bonus does not preclude anyone else's getting one also, fears of competition and failing to collaborate go away. At schools using the TAP, striving for the annual performance awards improves teaching and enhances collegiality and morale.

One cautionary note: Most school districts are constantly trying to get more money for their teachers. So some will agree to a performance pay plan, but then transform it into an across-the-board salary hike: defining high performance as things everyone does (taking classes toward an advanced degree), setting standards so low that everyone gets the maximum award, or giving all teachers the same evaluation. Districts must guard against such behaviors.

Can It Last?

Merit pay plans are expensive, especially if the performance awards are added to the salary schedule, so there are questions about whether the extra funds will continue to be available during the next economic downturn. What if the superintendent or school board turns over? Will results be demanded too quickly, and will the program be discontinued if test scores do not rise in a year or so?

Establishment of a dedicated funding source, such as an increase in the property tax levy or sales tax, could ensure sustainability. The former has been accomplished in Eagle County, Colorado, and the latter has the potential to support performance pay in the state of Arizona, where voters passed a law in 2000 to raise the sales tax by $0.006 and dedicated 40 percent of the proceeds to performance pay. In fact, most districts in Arizona ended up defining performance pay in ways that gave all teachers the same increases (for example, by allowing more course credits to fill the requirements for meritorious work). The district using the TAP was one of the few that actually used the money for real performance pay. The bottom line, though, is that support and advocacy by teachers is key to sustainability.

Although many teachers initially view performance pay and accountability as contentious issues in reform, absolute levels of acceptance for all the principles embodied in the TAP are high. Since the inception of the TAP, surveys of teachers' attitudes toward the elements of the program show that collegiality and teachers' satisfaction have remained strong in the schools using the TAP. This finding refutes many who argue that pay for performance leads to increased competition and divisiveness. These attitudinal results reflect the holistic approach of the TAP, which combines an accountability system having clear rewards and a professional development system to support all teachers (veteran and novice) in improving their class-room instruction.

Another important way to ensure sustainability is to show that the program is working. There is always the fear that results will be demanded too soon, and then the program will be discontinued if test scores do not rise in a year or so. We now have three years of results from TAP schools in Arizona and two years from TAP schools in South Carolina. We compared 25 year-to-year changes in student achievement in TAP schools to control schools. In 17 of these cases, or 68 percent of the time, the TAP schools outperformed their controls. This compares favorably with the results of a RAND evaluation of schools that have initiated other comprehensive school reform programs. RAND concluded that 50 percent of the schools with these reforms outperformed the control schools in math and 47 percent outperformed the control in reading, although these schools had been operating for a substantially longer period of time than the schools using the TAP. One important anecdotal explanation for the success of the TAP is that teachers in schools using the program improve significantly because their performance evaluations are related directly to TAP teaching rubrics.

Any pay-for-performance plan in K-12 education will succeed only if teachers buy into it from the start, if it is fair, and if it is embedded in a systemic reform that supports all aspects of performance reward, especially those that encourage teachers to become better at their craft. Such a pay plan will be revolutionary, but will not become obsolete in any sense of the word.

Patience as a Virtue (Figure 1)

Under the rigid salary structure of many public school systems, teachers
must work for more than 20 years before they can earn a $50,000 annual
salary.

Average Base Salary of U.S. Teachers in 1999-2000 by Years of Experience

                      Average Base Salary
Years of Experience   (current dollars)

First                 29,088
 2 to 4               30,434
 5 to 9               34,066
10 to 14              38,712
15 to 19              42,942
20 or more            48,115

SOURCE: National Center for Education Statistics

Note: Table made from bar graph.

Compared to the Private Sector (Figure 2)

Starting salaries for teachers are competitive when compared to college
graduates with a degree in English, but less so for those graduates who
specialized in a field that requires mathematical skills, such as
accounting or computer science. (Salary figures do not include
benefits.)

Average Base Salary Offered to College Graduates in 2004

                  Average Starting Salary
College Major     (current dollars)

Teaching          30,496*
Accounting        41,110
Computer Science  49,691
English           31,169

* This is the estimated beginning salary for the 2003-04 year as
provided in the AFT 2003 Survey and Analysis of Teacher Salary trends.
SOURCES: National Association of Colleges and Employers, American
Federation of Teachers

Note: Table made from bar graph.

BY LEWIS C. SOLMON

Lewis C. Solmon is executive vice president, education, at the Milken Family Foundation and director of its Teacher Advancement Program. He is the former dean of the Graduate School of Education at UCLA.