Errors are frustrating. Production line errors can lead to quality control rejects, with attendant production delays and increased costs. Clerical errors can introduce misinformation into the corporate MIS -- such misinformation is often difficult to detect, and time consuming to correct when it is discovered.

Error reduction policies based on frustration, however, are likely to cause more problems than they solve.

One example -- senior management at a client company, in an effort to deal with the problem instituted by a "Zero Error Tolerance" policy. It required each level of management to personally assure the complete accuracy of any and all information before passing it along to a higher level, and was enforced by varying degrees of disciplinary harshness.

In a sense the policy succeeded. Errors were reduced, if not to zero, at least to a very low level. That success, however, was costly. Company business slowed down as every process and report was repeatedly scrutinized, checked, and confirmed before any action was taken. Morale plummeted as well, both because of the police state atmosphere bred by the "Zero Error Tolerance," and the evident business stagnation.

Within a month the policy was rescinded.

This is not to suggest that error reduction is a futile goal; indeed, error reduction is both desirable and achievable. What's required, however, are the three "R's" of error reduction:

• Realistic Goals People are not error free. Policies need to recognize this fact.
  Driver training and warning lights help reduce grade crossing accidents; overpasses eliminate them.
• Representation Job changes are most effective when those whose working conditions are affected participate in designing the changes.
• Recognition Providing recognition of desired actions is a more powerful motivator than penalizing undesired ones.