Merit pay won't work, never has. Deming identified merit pay, annual reviews, etc. as one of the Seven Deadly Diseases of Management nearly 3 decades ago, yet folks like yourself hold to the idea that this time results will be different.

Taking action on the basis of results without theory of knowledge, without theory of variation, without knowledge about a system. Anything goes wrong, do something about it, overreacting; acting without knowledge, the effect is to make things worse. With the best of intentions and best efforts, managing by results is, in effect, exactly the same, as Dr. Myron Tribus put it, while driving your automobile, keeping your eye on the rear view mirror, what would happen? And that's what management by results is, keeping your eye on results.

Russ,

I read some of Deming’s work and philosophies and they are interesting. I myself believe that merit pay for teachers would be very difficult to implement in many ways agree with Steiny’s points.

But … I didn’t see anywhere that Deming advocated worker unions. I didn’t see where he advocated pay increases solely on experience. He also advocated personal responsibility and leadership of the workers. He advocated workers and management working together on shared goals. One of his other seven deadly diseases was excessive medical costs for the employer. Do you agree that is a problem too? Or just the merit raises.

In short, while Deming might agree that merit pay is not the solution, based on his philosophy I don’t think he’d agree with anything else you would advocate in terms of increasing production in any field.

In short, while Deming might agree that merit pay is not the solution, based on his philosophy I don’t think he’d agree with anything else you would advocate in terms of increasing production in any field.

msteven, a bit of a red herring considering we're talking about the "merits" of merit pay. And I'm not quite sure what to say about the personal dig other than to note that I make a decent living advocating total quality management best practices.

As to the rest of the comment, constancy of purpose has nothing to do with organized labor. TQM has been implemented in many unionized industries; there is nothing incompatible about the two. And let's not forget where Deming suggests we begin: "The problem is at the top; management is the problem."

(btw, you won't find many progressives who don't agree with Deming's 6th deadly disease)

You advocate TQM and I work as a statistician and am knowledgeable about use of SPC in production environments. Your quote about management being the problem may be accurate but that does mean not the worker has no responsibility. I’m sure Deming would agree. I’d go so far to say that often the problem with management is that they enable or continue to allow lack of productivity.

You have quoted Deming as saying that merit pay does not work and management is the problem. But as you say, we are talking about the “merits” of merit pay and you pointed out that Deming did not agree with it on the basis that it does not increase production. Does that you mean you advocate the current system of pay increases for teachers – on the basis that it does? If not, on what basis should there be pay increases for public teachers? This is what we are talking about.

"Does that you mean you advocate the current system of pay increases for teachers – on the basis that it does? If not, on what basis should there be pay increases for public teachers?"

I don't think seniority or education based pay is a bad solution, though not necessarily as a way of improving teacher/student performance. There's plenty of research that indicates pay is one least effective methods for improving employee performance.

Deming is said to have responded to the question of what to do instead by saying, do whatever Peter Scholtes says. Here's what Scholtes said (quoted at length this time):

There are two alternatives to performance appraisal that managers don't like to hear:
1. Change the way you think. Until managers let go of their obsession with the individual worker and understand the importance of systems and processes, they will not enter the quality era. Without this change in mind-set, managers will continue to look for alternatives that are no different from what they are trying to replace.
2. Just stop doing it. When you are doing something that is demonstrably harmful, you can stop doing it without finding an alternative way to harm yourself. Conventional managers are, in effect, beating their heads against the wall and asking, "If we stop beating our heads against the wall, what will we beat our heads against?"

One way to develop alternatives to performance appraisal is by debundling. If performance appraisal is a fragile cart asked to carry too heavy a load, we can remove each piece of baggage and build for each a separate vehicle designed specifically for that function. Here are some steps to take in debundling:
1. Acknowledge each service and expected benefit as something important for the organization to successfully accomplish.
2. Treat each as a separate function, disconnected from all the others.
3. For each service and expected benefit, ask: What is the best way to successfully accomplish this?
4. Set up separate systems and processes specifically designed to successfully accomplish that service or expected benefit.
5. When identifying the services and expected benefits and designing these systems and processes, keep in mind the principles at the heart of quality.

An example of a debundled function: giving each employee feedback on his or her performance. Why doesn't performance appraisal adequately provide an employee with useful feedback?
* Feedback is, by definition, something derived from one part of the system and given to an antecedent part of the system. Performance appraisal is a feed-down activity that is hierarchical by nature, not systemic. And the hierarchy is not ordinarily the best source of system-focused feedback.
* Performance appraisal, particularly when tied to income and promotability, engenders posing and pretense: easy targets, creative accounting, and evaluation that reflects the preconceptions of the supervisor and the manipulations of the person being evaluated. All of this works against the communication of data needed for improvement.

A better alternative is to have employees, alone or in natural work groups,
* Identify a key work process.
* Identify the customer(s) of that process.
* Learn what characteristics of the product or service--the output of the key process--are most important to the customer.
* Get feedback from the customer(s) on how well these characteristics are met.
* The next step is that managers must refrain from using this data to evaluate an employee and continue to encourage each employee to develop feedback loops such as this for all the key tasks.