The Effects of Managed Care and Prospective Payment on the Demand for Hospital Nurses: Evidence from California

Health Services Research, Dec, 1999 by Joanne Spetz

Third, I must develop measures of hospital technology. Including a measure of technological change is crucial in determining the demand for nursing labor because technology affects the demand for skilled labor relative to unskilled labor (Berman, Bound, and Griliches 1994). Prior studies of nursing employment have not included a measure of technological change.

In this analysis I use a Saidin index to control for changes in the technologies used by hospitals (Spetz 1995; Spetz and Baker 1999). This index is a weighted sum of indicators for various technologies and services, with the weights representing the percentage of hospitals that do not possess the technology or service. Rare technologies--rare because they are new, expensive, or difficult to implement--receive higher weights. Common technologies such as operating rooms receive low weights. A Saidin index was developed using weights from the 1981 data. [5]

The fourth concern is measuring the impact of selective contracting. There is no reliable source for city-level HMO and PPO enrollment data before the late 1980s. However, one can determine the percentage of hospital discharges for which the expected source of payment is an HMO from the OSHPD Patient Discharge Data. [6] I calculate the share of county non-Kaiser discharges insured by HMOs to measure the penetration of HMOs in local markets. [7] Although the effect on nursing employment of the share of an individual hospital's patients insured by HMOs is of interest, the relationship between these is probably endogenous. It is less likely that regional HMO penetration is endogenous with an individual hospital's staffing decisions. The share of discharges reimbursed by HMOs is likely to be a lower bound for HMO enrollment in a county because HMO patients typically have fewer hospital stays than non-HMO patients (Luft 1981). Before 1983's selective contracting legislation, there were virtually no non-Kaiser HMO enrollees; thus, I assume that the percentage of discharges reimbursed by HMOs was constant from 1976 through 1983.

HMO penetration does not explicitly measure the financial pressure placed on hospitals by managed care. HMOs bargain with hospitals individually or with multihospital corporations. Each contract between an HMO and a hospital produces a different level of financial pressure for that hospital. Thus, heterogeneity in the financial pressure caused by HMOs, either across hospitals or over time, is not examined in this study.

No explicit way exists to measure the growth of PPOs, even though they might be placing significant financial pressure on hospitals. To the extent that markets in which HMOs have greater influence are more conducive to competitive healthcare financing, HMO penetration might proxy for PPO penetration. The yearly dummy variables included in the demand equations also might measure some of the effect of both HMO and PPO growth.

Measuring the impact of the prospective payment system is my fifth concern. A measure of the financial pressure placed on hospitals by the prospective payment system is used to identify the effect of PPS. "Biteshare" is the expected reduction in revenue caused by PPS as a percentage of projected pre-PPS revenue. Gruber (1994) and Staiger and Gaumer (1992) used similar measures to evaluate the effects of PPS in studies of charity care and Medicare mortality.