Using Negative Reinforcement

So far we have considered variations on the theme of positive reinforcement. Negative reinforcement also has many uses. One is to increase productivity in industry. Most people will work extra hard if they can get some time off from work as a reward. This is negative reinforcement because the level of work is increased (the behavior becomes more frequent) as the result of having an aversive stimulus removed (time off from work). Sometimes this works too well!

Here is a story told by a guest lecturer in a Behavior Modification class.

How does the story about the automatic transmission assembly line illustrate the potential power of negative reinforcement?

A major automobile manufacturer asked some behavioral psychologists to help solve a morale and production problem at an automobile assembly plant in Ypsilanti, Michigan. Workers on one of the lines were going too slow, holding up the entire plant. They were supposed to manufacture 60 transmissions per hour, but they seldom achieved this objective. When managers urged them to speed up, they complained about being exploited, because (as they saw it) they were being asked to produce more work for the same salary. They felt abused. The supervisors were disgusted. Bad feelings existed all around.

The psychologists suggested a contingency. If the workers produced their quota of 60 transmissions before the end of the hour, they could take a break until the end of the hour. The psychologists conceived of this as a negative reinforcement contingency. They arrived at the idea when they realized there was no positive reinforcer available, because the management would not allow any extra salary incentives. OK, said the psychologists, then we will use a negative reinforcer. If the workers hate being "pushed" all the time, we will let them take time off (and avoid being pushed) when they meet their quota.

The program worked like magic. Productivity leaped. 60 transmissions were produced in 50 minutes, leaving 10 minutes for a break every hour. The supervisors felt strange seeing the workers "goof off" every hour, but at least the line was finally meeting its quota. Then a funny thing happened. The line began to manufacture 60 transmissions in 45 minutes, then 60 transmissions in 40 minutes. Soon workers on the other lines were grumbling, "Hey, how come those guys are getting a 20 minute break every hour?"

The plant managers had not expected this to happen, and they had no plans for dealing with it. They boosted the production quota to 80 transmissions per hour. The workers grumbled, but given the alternative of going back to the old plan with no hourly break, they accepted the new quota. Nobody expected the "problem" that occurred next. At first, producing 80 transmissions took almost the whole hour. But soon 80 transmissions took only 50 minutes...and then only 45 minutes, and the workers were back to taking a 15 minute break every hour. Workers on the other assembly lines started asking for a similar system.

At this point, the plant managers decided the experiment had to end. It was unacceptable to have people taking a break every hour. The managers sent the behavior modifiers home and went back to the old system. "We knew it wouldn't work," the managers said. [Author's files]

The assembly line story is an example of negative reinforcement. The reinforcer (time off from work) involved removal of an aversive stimulus so it was "negative" reinforcement. Like positive reinforcement, it produced a higher frequency of behavior...until the managers refused to let it continue.

Eventually a simplified form of this incentive did find a home in the auto industry, and it even became part of union contracts. In an echo of the above story (which I heard as an undergraduate in the early 1970s) news articles in 1997 reported that workers in a stamping plant at Flint, Michigan were going home after only 4 hours instead of the usual 8. Apparently the union had negotiated a production quota based on the assumption that certain metal-stamping machines were capable of stamping out 5 parts per hour, but actually the machines were capable of 10 per hour. The workers speeded them up to 10 parts per hour and met the quota specified in their union contract within 4 hours. GM decided to eliminate the "go home early" policy, and this was one issue in a 1998 strike against General Motors. In the end, a modified version of the policy with a higher rate of production was re-affirmed in a new contract.

What is reinforced by an hourly wage? By piecework? What is the disadvantage of piecework?

If you think about it, an hourly wage reinforces slow behavior. The less energy a worker puts into the job, the more money the worker receives per unit of energy expended. By contrast, when people are paid according to how much they produce (so-called "piecework" systems) they work very quickly to maximize their gain. The obvious disadvantage is that workers who are rushing to complete their quota might produce a poor quality product, or endanger themselves, unless steps are taken to maintain quality control and safety on the job.

Babies as master behavior modifiers

Babies are masters of behavior modification who use negative reinforcement to increase the frequency of parenting behaviors in adults. Non-human babies of all different types practice this form of operant conditioning instictively. For example, juvenile birds fresh out of the nest will fly from branch to branch, following their parents, making unpleasant noises until fed.

How are babies "masters of behavior modification"?

Almost nothing is more aversive to parents than the cry of a baby, and adult humans will do almost anything to eliminate or prevent that stimulus. They will feed a baby, change its diapers, dance around with it, get up in the middle of a deep sleep…all to prevent crying. Fortunately for parents, crying (in babies under one year old) is not reinforced by the application of love.

What finding surprised behavior modifiers?

This finding actally surprised behavior modifiers when it was first discovered, because it was so counter-intuitive. Studies showed that babies whose parents responded less to their crying (leaving the baby crying longer or more often) actually suffered more crying in the future. Perhaps this is because crying cannot be completely extinguished. Parents must respond eventually, so parents who respond slowly are essentially using intermittent reinforcement and are teaching their babies persistence, not quietude.


Write to Dr. Dewey at psywww@gmail.com.

Don't see what you need? Psych Web has over 1,000 pages, so it may be elsewhere on the site. Do a site-specific Google search using the box below.

Custom Search

Copyright © 2007 Russ Dewey