Nursing personnel constitute over half of hospital workers, yet surprisingly little is known about differences in nurse staffing across hospitals.

This study analyses hospital demand for nursing staff using a cross sectional sample of 524 hospitals from the American Hospital Association's (AHA) Nursing Personnel Survey (1989). The AHA data are unique because 1989 was the first year nursing personnel were identified by specialty area. Thus, it was possible to examine nurse employment on medical surgical units where most of the nursing substitution takes place.

Regression analysis was used to estimate separate wage and employment equations for registered nurses (RN), licensed practical nurses (LPN), and nursing assistants (NA). Cross-sectional coefficients were then used to see how well the model could predict employment changes during the eighties. The model accurately predicted drops in LPN and NA employment, but could not show the expected increase in RN employment.

The findings further indicated a marked lack of variation in RN wage across hospitals. Some monopsony influence was noted, yet, wages were found to be determined primarily in regional and hospital markets. Although greater employment was associated with higher hospital output, proportional use of staff decreased suggesting hospitals were economizing on their use of nursing personnel. Employment estimates also indicated substitution between RNs and NAs, but the expected pattern of substitution between RNs and LPNs was not observed.

Limitations to the study include measurement of technology which is hard to observe and always changing, and the inability to account for supply side factors. In particular, better work alternatives for women may have released crowding in female dominated occupations, and reduced hospital monopsony power in the RN market.

Key policy considerations for nurses, hospital employers, and policy makers include: Low RN relative wage will encourage hospitals to hire all the available nurses and foster inefficient use of RNs. Higher RN wage, however, would eventually translate into fewer jobs, so restructuring work and adjusting state practice acts must be considered. Also, claims of increasing nursing skill mix need to be examined, particularly in view of increased employment of less trained two-year RNs. Finally, the interplay of social and economic forces should continue to promote better work alternatives for women. Although there are benefits to increasing RN supply through shortened education programs, policy makers should be wary of deskilling nurses, the backbone of the hospital industry.