How Does "Physician-Induced Demand" Affect Individual Physicians' Cesarean Rates?

By Jill Arnold

As reported yesterday on the U.S. News & World Report Economic Intelligence blog by Sita Slalov, researchers Erin M. Johnson and M. Marit Rehavi examined the effect of physician-induced demand by comparing "physician-mothers" with their non-physician mother counterparts whose first birth occurred in California between 1999 and 2005. In a nutshell, the study looked at whether a patient’s medical knowledge affects treatment and outcomes. After controlling for demographics, risk factors and hospital type, they found that physician moms were 9% less likely to have unscheduled cesareans and had reduced morbidity. This is attributed to physician-moms’ medical knowledge making them inherently informed patients, which in turn makes them less impacted by the financial incentives of their treating physician. It remains consistent with the hypothesis that physicians are able to avoid over-treatment and unnecessary medical procedures in their own care.


Slalov highlighted that some hospitals, such as those which are HMO-owned, do not incentivize unnecessary cesareans because doctors are salaried. Slalov wrote:

Johnson and Rehavi found that non-physician mothers who gave birth in non-HMO-owned hospitals were indeed more likely to have C-sections than non-physician mothers with similar characteristics who gave birth in HMO-owned hospitals. That supports the claim that doctors' financial incentives play at least some role in their treatment of uninformed patients. But there's even stronger evidence for the physician-induced demand hypothesis: it turns out that physician-mothers were about equally likely to have C-sections regardless of whether they gave birth at an HMO-owned hospital. In other words, financial incentives don't seem to affect the choice of treatment for informed patients.


Johnson and Rehavi's paper, Physicians Treating Physicians: Information and Incentives in Childbirth, is a working paper and has not yet been peer-reviewed. 

In their paper, the authors define the physician-induced demand hypothesis, which puts forward that because patients do not have the necessary medical knowledge to make treatment decisions, and because doctors profit from recommended treatments, doctors will therefore shift patient demand and “move treatment quantity in the direction of their own preferences.”

Two paragraphs of the paper are dedicated to the relationship between physician income/reimbursement and c-section rates. 

Numerous authors have documented a positive cross-sectional correlation between physician supply and rates of surgery (Fuchs (1978), Cromwell and Mitchell (1986), Rossiter and Wilensky (1983)). Following Dranove and Wehner's (1994) critique, this empirical approach was superseded by studies exploiting exogenous shocks to physician incomes.5 Gruber and Owings (1996) provides credible evidence of PID by exploiting the shock to obstetrician incomes resulting from the secular decline in fertility rates in the 1970s. They found that a 5% fall in incomes leads physicians to increase the C-section rate by 1 percentage point.

A related test for inducement exploits changes in physician fees. In response to a fee reduction, physicians have been found to make up lost revenue by increasing volume (Nguyen and Derrick (1997), Yip (1998), Jacobson et al. (2010)).6 There are also studies which found little evidence of income effects, with physicians altering quantities in the direction of the fee change. For example, Gruber et al. (1999) finds an increase in the C-section rate in the Medicare population after C-sections became more highly reimbursed relative to vaginal deliveries. Specifically, they found a 0.7 ppt increase for a $100 increase in the fee differential. In both of these approaches identification comes from shocks to providers; as a result they cannot estimate the overall level of PID.

Although provided as background for their research, the previous literature cited makes it appear that the effect on volume (i.e., the 0.7 or 1 percentage point increase in cesarean rate) would be minimal.