Academic Articles Health Insurance Public Policy

The Effects of HMO on the Survival of Hospital Services

In the mid to late 1990s, there was a large backlash against HMOs due to the perception that the “HMO cares more about saving money than it does about them.” People feared that HMO would drop coverage of services which they valued.

In this month’s NBER working paper edition, Yu-Chu Shen (2006) looks answer two questions related to this topic:

  1. Does the degree of HMO penetration in a MSA increase the likelihood that a hospital will drop some of its services?
  2. Does the percentage of for-profit HMO enrollment increase the likelihood that a hospital will drop some of its services?

Shen divides hospital services into two groups. The first consist of safety net services such as having an emergency department, a trauma center, HIV/AIDS services, and substance abuse services. The second group of services are safety net services which are generally profitable. Examples of this group are maternity care, birthing rooms, child wellness services, women health centers and sports medicine.

Using a hazard function (specifically the Cox proportional hazard model), Shen tests his hypothesis using dummies for low, medium and high HMO penetration and then repeats the procedure for low, medium and high dummies for the percentage of HMO enrollment which have for-profit HMOs.

In his results, Shen does not find any consistent trend in the abandonment of services, with the exception that in the 2000-2003 period it seems that areas with high level of HMO penetration dropped profitable services less frequently than low HMO penetration areas. Social pressure may be the reason there were few major difference in the hazard rates of dropping unprofitable safety net services between Shen’s categories. The study did find that government ownership is associated with a lower hazard of shutting down unprofitable services. Large hospitals and teaching hospitals have a lower hazard rate of shutting down any service.