The answer: more competition leads to more technology adoption. This is the finding from a study by Karaca-Mandic et al. (2016). They use data from 100% Medicare claim in 2003 and 2004 as well as linked information on hospitals [American Hospital Association (AHA) Annual Survey] and physicians [American Medical Association (AMA) Masterfile].
Competition is measured two ways. First, the authors the traditional Herfindahl– Hirschman Index (HHI), which is the sum of the squares of the firms market share. Although one can easily measure HHI within a given area, HHI does not account for the location of patients within a given area. The authors use a second approach, where the measure market share for each practice among patients who lived within 50 miles. Following the approach of Berry (1994), they run a conditional logit at the patient level where the dependent variable was ln(shjk) – ln(sh0k), where shjk was the market share of practice j in ZIP k (the proportion of patients in ZIP k choosing practice j), and
sh0k was the share of patients living in ZIP k who chose a practice outside of the choice set (beyond the 50 mile radius). The explanatory variables were the distance from the patient’s home ZIP to the ZIP of the cardiologist’s office, this distance squared, and a vector of practice fixed effects.
The study calculates measures of competition for both physician practices and hospitals. Using either measure of market competition, the authors find that:
…substantial variation in use existed across cardiologists and hospitals. We found significant evidence that the structure of the cardiology practice market mattered. [Drug eluting stents] DESs diffused faster in markets where cardiology practices faced more competition. This finding supports the hypothesis that competitive pressure to maintain or expand PCI volume shares compelled cardiologists to adopt DESs more quickly.
On the other hand, hospital competition was not a determining factor:
“…we found no obvious evidence that the structure of the hospital market mattered.We conjectured that hospitals faced additional, conflicting influences that did not exist at the cardiologist level. Hospitals would have faced significant pressure to secure PCI volume through adoption, but they also faced the direct cost of purchasing DES.
Although this study is interesting, would its findings hold up today? Has hospitals are buying physician practices and physicians are banding together to form accountable care organizations, the line between physician, hospital and insurer are starting to blur. For these new organizational forms, will competition be a key driver of innovation adoption–as it was for physicians and drug eluting stents–or will it be a small factor–as it was for hospitals in this case. That question, has yet to be resolved.
- Karaca-Mandic, P., Town, R. J. and Wilcock, A. (2016), The Effect of Physician and Hospital Market Structure on Medical Technology Diffusion. Health Services Research. doi: 10.1111/1475-6773.12506
- Berry, Steven T. “Estimating discrete-choice models of product differentiation.” The RAND Journal of Economics (1994): 242-262.