Deere's corporate efforts to manage health care costs go back to the late 1960s and early 1970s. The company's annual health care bill as recently as 1972 was $20 million. At that time it decided to self-insure and to self-administer health and accident claims of employees, dependents, and retirees. By self-administering its claims, the company built a data base on utilization and costs -- information that companies were not generally getting from their insurers. Deere was one of the earliest corporations to become self-insured (a recent survey of 500 firms found that nearly 40 percent of them were now at least partially self-insured).
Having established an efficient claims payment mechanism, Deere came to realize that it nevertheless remained powerless to control the ever-rising health care bills fed into the mechanism. Selfinsurance and self-administration produced only administrative cost savings, which -- though nice to have -- were minuscule compared with the price tag for delivery of services to its employees. By 1977 the company's health care bill had climbed to $60 million, and it decided to conduct a thorough study of how it was managing its health care costs.
The study prompted two major decisions: (1) the company should reorganize itself to give sufficient management attention to the task; (2) after putting its own house in order, it should try to act jointly with providers in dealing with health care issues.
Deere's first step was to form a new health care department that brought together various functions concerned with health care that were scattered around the company. By creating a high-level position for the manager of this department, the company made plain that it was serious about doing more than just studying the problem. This position was recently raised to a higher managerial level, indicating Deere's commitment to gaining greater control of its costs by managing rather than merely administering the health care function.
Deere also created a health care systems group to manage its claims processing and develop the claims data base needed to try to