Propper C, Burgess S & Gossage D, (2003)

‘Competition and Quality: Evidence from the NHS Internal Market 1991-1999’

CMPO Working Paper, University of Bristol

Estimates the effect of competition (as introduced by the NHS Internal Market reforms) on quality of care in NHS hospitals.
Quality is measured using the death rate from acute myocardial infarction (AMI).
Level of competition is measured in 3 ways, each relating in some way to the number of hospitals within 30 minutes drive of the hospital under consideration.
Uses panel data from 1991 to 1999 to exploit the fact that competition policy changed over the 1990’s – it was introduced in 1991, actively promoted until 1995, downplayed from 1996 until 1997, then actively discouraged post-1997.
The panel data and analysis allows more detailed analysis and conclusions than the authors’ similar (2000) paper.

Key results:

The authors find evidence that the effect of competition was to reduce quality. Hospitals located in more competitive areas have higher death rates during the period.
The estimated effect of competition is robust to different measures of competition and hospital volume.
The disparity in death rates between the most and least competitive areas is greatest around 1995, when competition was at its height.
Cumulated over the entire period, the higher death rates associated with competition more than offset the positive impact of technological change in reducing death rates.
However, if hospitals competed by simultaneously reducing quality for services (like AMI) with low demand elasticity (AMI is an Accident & Emergency condition) and increasing quality where demand elasticity was higher, it is at least possible that competition could have raised overall quality.