A good doctor is hard to find—and for those with Medicaid or without health insurance at all, finding a new primary care doctor is almost impossible. According to an audit study headed by physician Karin Rhodes, the difference in access to primary care is even more extreme than we might expect. New patients with Medicaid were far less likely to obtain a doctor’s appointment than their privately insured peers, and uninsured patients had it hardest of all: only 15.4% could obtain an appointment without paying more than $75 at the time of service.

One reason this disparity has gone undetected until now is that physicians overestimate how many Medicaid patients they treat by up to 40%. In order to get a true read, Rhodes uses an audit study, which is perhaps the most powerful tool social scientists can use to measure discrimination. In Rhodes’ simulated patient study, a team of 10 field staff members was selected for diversity in race and age based on the sound of their voices in a phone audition. After being trained to pose as new patients, they made 11,347 calls to doctors’ offices in 10 states to assess “business as usual”. On each attempt to make an appointment with a primary care physician, a caller was prompted to adopt the persona of a patient with one of three insurance types: private, Medicaid, or no insurance. In this way, two otherwise identical “patients” were presented in the real-world situation of making an appointment, and the only characteristic that varied was their insurance type, which was experimentally manipulated by the researcher. As a result, using the audit method allowed Rhodes’ team to test exactly how much of the difference in outcomes was due solely to insurance type.

Overall, 87.4% of privately insured callers were able to schedule an appointment, compared to 57.9% of Medicaid callers. Among uninsured patients, 78.8% were able to see the doctor, but only if they could pay a fee of $75 or more in full at the time of service. The median out-of-pocket cost for a primary care visit was $120, and fewer than one-fifth of practices allowed flexible payment arrangements.

Rhodes’ study assessed the capacity of the primary care system before the Affordable Care Act’s 2014 coverage expansion. The ACA is projected to cover 25 million formerly uninsured Americans. Because access to primary care is considered vital to improving population health outcomes, system that is already strained may make the ACA’s goals harder to achieve. This means that the system of providers accepting Medicaid needs to be strengthened before coverage increases will translate into gains in access to primary care.