Confusion over merit pay sometimes occurs when union groups use the term to describe “differential pay,” which offers higher compensation to encourage teachers to take jobs in high-needs schools and high-needs subject areas. Differential compensation programs are distinct from merit pay in that differential pay rewards teachers and principals for their decisions to work in settings with chronic vacancies, not for classroom performance per se. Differential compensation can come in the form of higher pay or in the form of loan forgiveness, signing bonuses, housing subsidies or other incentives.

Differential pay is commonplace in other professional settings, including higher education. Liberal arts professors, for example routinely earn less than professors in business departments. Within business departments, economics professors regularly earn less than finance professors.

Two basic economics concepts — substitutes and supply — explain these practices. The concept of substitutes acknowledges that business professors have more higher-paying job opportunities outside of teaching than do liberal arts professors. The concept of supply recognizes that when the labor supply of liberal arts professors exceeds the demand represented by professorships and substitute professions, wages will stay low.

In primary and secondary education, differential compensation means that teachers willing and able to fill vacancies in high-needs subjects or high-needs geographical locations would receive a better compensation package than colleagues of similar seniority, credentials and accomplishment. In Michigan, as in many states nationwide, schools in certain rural and urban areas have difficulty filling positions in special education, math, science and foreign languages.[*] Research also suggests that small rural and urban schools serving disadvantaged students tend to be staffed by less experienced and academically weaker teachers.[167] To address supply-and-demand problems resulting from geography, a compensation system including differential pay components might offer a salary supplement to attract and retain teachers who have either proven their effectiveness or who have the potential to be effective.

It is not uncommon for teachers to gain a few years of experience in a more challenging or remote setting only to leave for easier or less remote settings as soon as the opportunity arises. The American Federation of Teachers agrees that differential pay could help address teacher shortages in hard-to-staff areas and hard-to-staff fields.[168] Researchers who tend to disagree with many of the policies advocated by unions also support differential compensation.[169] For example, the American Enterprise Institute’s Fredrick Hess has succinctly written, “Beyond teacher effectiveness, however it is measured, there are several other considerations that districts should acknowledge and compensate: The relative challenges an educator faces, the desirability of the work environment, and the relative scarcity of the teacher’s skills.”[†][170]

One unresolved question concerning differential compensation is just how large salary incentives need to be to attract teachers into hard-to-staff schools. Researchers for the School Finance Redesign Project at the University of Washington recently released a study that attempted to estimate the appropriate amount teachers should be offered to work in schools serving disadvantaged students.[171] The authors established that a teacher in the private-school labor market is paid more to work in a disadvantaged school, and they suggest that the differential pay in private schools might be a starting point for calculating “combat pay” for public schools. However, the authors noted serious technical limitations in modeling such wages and were unable to arrive at an estimate.[**] Similar problems face other researchers trying to estimate exactly what should be offered as a bonus or salary supplement. As a practical matter, these problems simply mean that it is currently difficult to predict just what levels of differential pay will be needed; policymakers will have to experiment with various levels and types of differential pay to discover what works.

Nevertheless, education researchers Robert Gordon, Thomas Kane and Douglas Staiger, writing for the Brookings Institution’s Hamilton Project, have offered the suggestion that teachers in the top quartile of teacher effectiveness should be offered annual bonuses of at least $15,000 to work in high-poverty schools.[‡] Such an incentive would require substantial funding increases, which some suggest should be supplied by the federal government.

Ideally, instead of seeking more federal involvement in Michigan’s public schools, state and local policymakers might try reconceptualizing how salaries are determined. To increase salaries for certain teachers will require either more discretionary monies for teacher compensation — a tall order in a period of rapidly rising fixed pension and employee health care costs[172] — or a redistribution of existing funds within or across budget items. For example, if a local district were to scrap the single-salary system and pool all the available money for teacher compensation, it could then allocate to each school its appropriate share. Then a principal could pay teachers in different subjects and grade levels the amount necessary to fill his or her staff with the highest-quality personnel he or she could afford.[***]

Such a scenario would no doubt be a tough sell during collective bargaining negotiations. Still, it would help to address the fact that most schools are currently overpaying some teachers and underpaying others, such as their physics teachers (if they can find some). Although this observation may sound like a value judgment about the inherent worth of physics compared to other subjects, it is not. Rather, it is an assertion that physics teachers are often underpaid because there tends to be a small supply of them relative to the number of teaching positions available and relative to high demand for their skills outside the teaching profession. If teacher compensation were decoupled from the single salary schedule, a principal could likely afford a physics teacher at a genuinely competitive salary and yet still afford a good, but traditionally overpaid teacher in another subject at a more reasonable rate.[††]

Instituting such a system would of course have practical problems, but many of those concerns could be addressed creatively. Though a school district might be wary of challenging all current compensation norms and expectations for existing teachers, it might try bargaining with the unions to allow the hiring of new teachers under a reformed compensation system that would grant principals the discretion to pay differentially.

Fully implementing differential pay across different subject areas or grade levels would probably require abolishing the single salary schedule. Although this suggestion may appear politically infeasible at present, differential pay is quite common in private schools, where teachers with relatively similar experience and credentials often earn quite different salaries. Private-school teachers are sometimes grouped according to experience — zero to five years of experience, six to 10 years of experience, 11 to 15 years of experience etc. — and principals can negotiate salaries with individual teachers based on the school’s staffing needs. So, if the annual salaries of teachers in the first band range from $30,000 to $50,000, a principal can offer a new physics teacher $40,000 and a new physical education teacher $30,000. These teachers could receive annual performance raises and still fall within the salary band.

Such an approach may be particularly necessary in urban and rural school districts that have a harder time filling certain positions where the supply of teachers may not be plentiful. Some of these districts already receive supplemental funding based on their count of at-risk students or their geographical isolation,[173] but to the extent this money was found to be inadequate, money for differential pay could be drawn from budgets for professional development (which is generally ineffective; see “Limited Role of Professional Development”).

Ultimately, the single salary schedule need not be considered sacrosanct, and better options, tailored to address the staffing needs in a local district, can reasonably be brought to the negotiating table. In addition to redistributing money across or within line items at the state and local levels, local districts and schools can also sometimes find support for differential pay incentives from private foundations, which have already demonstrated willingness to support such plans. In Michigan, one creative plan already exists to differentially compensate teachers willing to work in Detroit. The MDE reports:

The Eli Broad Foundation currently has two simultaneous programs operating in Michigan. Part of this program includes the recruitment of high school students from Detroit Public Schools (DPS) to attend Michigan State University (MSU) with full tuition coverage. This program requires a five-year commitment to return to teach, as a Highly Qualified teacher, in DPS. The companion to this program brings “Broad Fellows” (MSU students) into the DPS system over the summer months to assist in providing supplemental instruction to underachieving students. The Broad Foundation has given $6,000,000 to underwrite the cost of these programs in Michigan.[174]

While we have observed that there is support for differential pay across the political spectrum, policymakers should know that they will probably not pass such reforms without resistance. Put to the test, unions might choose to stand by the single salary schedule. Moreover, it is one thing for union members to support the idea of paying someone an extra $5,000 to work in an inner-city or remote rural school; it is quite another to institute a policy that would pay a history teacher and a calculus teacher with the same experience within the same building considerably different salaries. Nevertheless, a true differential pay system that can fill vacancies in hard-to-staff positions needs to do just that.


[*] In the “NCLB Revised Highly Qualified Teacher State Plan,” the Michigan Department of Education states, “There continues to be a statewide shortage of special education teachers as well as a need to improve science (specifically chemistry and physics) and math instruction.” The NEA reports these shortages for Michigan in “NEA Student Program.”

[†] As Dan Goldhaber notes, “Private sector compensation, in contrast, generally reflects not only individual attributes (often including an individual’s performance on the job) but also the attributes of a particular job.” (See Goldhaber, “Teacher Pay Reforms,” 7.)

[**] Dan Goldhaber, Kate Destler and Dan Player, “Teacher Labor Markets and the Perils of Using Hedonics to Estimate Compensating Differentials in the Public Sector” (Center on Reinventing Public Education, 2007), www.crpe.org/cs/crpe/download/csr_files/ wp_sfrp17_goldhaber_aug07.pdf (accessed May 19, 2008). The authors discuss the assumptions of hedonic modeling and the potential limitations of its use in estimating compensating differentials in public school teaching. The first assumption of this technique is that “models must completely account for teacher quality, otherwise the resulting wage premiums will mask quality differentials.” They note that the problem with the first assumption is that there are some difficulties in measuring teacher quality. The second assumption is “that hedonic wage models implicitly assume that wages are reasonably flexible so that they equilibrate the supply and demand for various teacher and school attributes and this assumption is unlikely to hold, as school districts and schools do not operate within a fully competitive labor market.”

[‡] Gordon, Kane and Staiger, “Identifying Effective Teachers Using Performance on the Job,” 19. They also pointed to the recommendations of other researchers about the appropriate amounts of differential pay that would be necessary to affect the teacher labor market in a way that would improve aggregate teacher quality and sorting. They wrote: “There is no settled answer to the question of how large incentives must be to attract and retain high-quality teachers in low-performing schools. Kate Walsh (2005) of the National Council on Teacher Quality suggests that bonuses would need to be 10 to 20 percent of base pay. Others have suggested that even 15 percent is inadequate (Miller 2003), that bonuses would need to be at least $20,000 to have an impact (Rothstein 2004), or that bonuses would need to range between 20 and 50 percent of base salary to attract teachers to the highest-poverty schools (Hanushek, Kain, and Rivkin 2001).” Goldhaber cites another study: “In their 2006 paper, ‘Would Higher Salaries Keep Teachers in High-Poverty Schools? Evidence from a Policy Intervention in North Carolina,’ Charles Clotfelter … and his colleagues study the North Carolina Public Schools Bonus Program (for the period from 2001-04), which awarded $1,800 annually to teachers in hard-to-staff subjects and schools. They find that this amount was enough to reduce turnover rates by roughly 12 percent.” See Goldhaber, “Teacher Pay Reforms,” 17.

[***] By contrast, Goldhaber recommends reforming teacher pay from the state level because “States, unlike localities, have the capacity to develop data and analysis systems that can credibly be used to assess significant areas of shortage, track teacher performance, and/or administer a differentiated pay system. And from a political perspective, it may be necessary to get to the state level in order to buffer some of the negative local political consequences arising from various pay reforms.” See Goldhaber, “Teacher Pay Reforms,” 25.

[††] For an excellent discussion of the economics of teacher supply, see Loeb and Reininger, “Public Policy and Teacher Labor Markets: What We Know and Why It Matters.” A similar point is made by Chester E. Finn, “Too Many Teachers, Too Little Pay,” (Hoover Institution, 2005), www.hoover.org/publications/digest/2993151.html (accessed May 19, 2008).