Features

Is the Private Sector a Model for Merit Pay?

by Richard Rothstein

It has become commonplace for education policymakers to argue that the way to improve teacher quality is to pay teachers more if they are more effective. This, the policymakers say, is how the private sector works.

But it is hard to find private-sector examples for such merit pay proposals. True, an exception is stockbrokers and sales clerks who are paid on commission. But we all know this system can engender high-pressure tactics that benefit the commission-paid workers at the expense of their customers. Used-car salespeople may be an extreme case, but recent prosecutions of Wall Street brokers show that conflicts between the interests of incentive workers and their clients are always potentially present and must be carefully monitored.

Similar problems would arise in schools, where merit pay systems, when based heavily on test scores, give teachers improper incentives to emphasize tested over equally important untested aspects of the curriculum. And do we want school children to be pressured in ways that might be effective in getting adults to buy a used car?

Tying Outcomes
When quality of work is important, corporations do not generally evaluate college-educated employees by numerical goals. For most professional jobs, it is hard to isolate the contribution of a single employee. Quantitative criteria are rare, even for group goals.

Consider management consulting. The pay of consultants cannot reasonably be tied to numerical outcome goals, such as gains in a client firm's stock price relative to overall market indices. It would be impossible to determine what role consultants' advice played in a company's success or failure. Also, stock prices can rise and fall. A collapse this year might result from bad advice given five years ago when the price was rising. Should the compensation of consultants who gave advice to Silicon Valley dot.com start-ups be based on skyrocketing stock prices in 1999 or on worthless equities in 2002?

Should a 4th-grade teacher be rewarded for better scores of her students this year, rather than basing her pay on whether 8th-grade scores of her students collapse, four years later, because a solid foundation for future success was sacrificed for short-term compensable gains?

Such difficulties in tying professional work to short-term measurable outcomes mean that, for professionals in the private sector, decisions on merit pay are mostly subjective. Firms typically base merit pay on qualitative, not quantitative, evaluations, relying upon input from supervisors, co-workers and even subordinates.

When it comes to merit pay, the practices of many retail firms toward their professional employees also do not fit the stereotype of pay-for-quantity alleged by those who want schools to adopt performance-based measures. Wal-Mart, for example, asks every managerial and administrative employee to negotiate annual goals with a supervisor. Raises are then based on management's judgment of how the employee related to customers and to co-workers and on management's subjective evaluation of work quality, recognizing that even good work can have poor results because of factors beyond an employee's individual control.

Pay for performance in private schools also does not resemble ideas being floated for public education. Certainly, elite academies do not use test data to calculate teacher pay. Nor, typically, do private schools for the less privileged.

Even for-profit education firms don't use merit pay in a way that many education policymakers imply. The Edison Schools, for example, does not tie individual pay directly to test scores. Edison, however, does give schoolwide bonuses, based not only on test scores but also on measures of parent satisfaction and supervisory evaluations of programs in music, art and other difficult-to-test outcomes.

For individual raises, Edison principals recommend small merit raises based on classroom observations, judgments about how well teachers work in teams, how they involve parents and the quality of student work portfolios, including test scores for reading and math, where appropriate. Such increases average only about 3 percent of pay.

Costly Oversight
An ironic aspect of the demand for teacher merit pay is that those who advocate it often also claim schools have too much bureaucracy, with too much money being spent outside the classroom. Yet a workable merit pay system requires much more administration than schools now can afford.

In the private sector, when pay is based on merit, managers devote many hours, sometimes as much as a week, to each employee's evaluation. Young law firm associates have senior partners for mentoring, supervision and modeling, where the partner-to-associate ratio is typically one to one. Business executives rarely have more than five subordinates reporting to them. At newspapers, editors often supervise no more than 10 reporters. But in public schools, it is common for one principal to oversee 25 teachers, often with little or no administrative assistance. No merit system that is based on careful supervisory evaluation, relying heavily on qualitative performance measures, can succeed with a 1:25 supervisory ratio.

Perhaps pay for performance could help improve public schools. But if the private sector is the model, designing merit pay requires more care than yet seems evident.

Richard Rothstein is a research associate of the Economic Policy Institute and a visiting professor at Teachers College, Columbia University. He is the author of Class and Schools (Teachers College Press). E-mail: rr2159@columbia.edu