This 2012 paper by Meliyanni Johar (University of Technology Sydney), published in the journal Economics Letters (ungated version here), investigates the question of whether doctors charge higher income patients more (for a standard general practice (GP) consultation). Higher income patients might be expected to have more inelastic demand for healthcare (because the cost of a visit to the doctor would take up a lower proportion of their income). So, given that the cost of providing a consultation is likely to be the same for high-income and low-income patients, if there is an observed difference in price between these two groups, it is an example of price discrimination.
In the paper, Johar uses a large dataset (2.3 million consultation visits, for 267,000 patients in New South Wales). She finds:
As expected, doctors charge higher fees to high income patients, but contrary to the hypothesis that fee is an indicator of quality, high quality doctors charge lower average fees than low quality doctors. The differences in average fee gaps however are small in size and not statistically significant. On average, there is a fee gap of about $6 for all GPs and $9–$10 excluding the 100% bulk-billing GPs. The latter is about 25% of the floor price, which may be regarded as large, for a basic consultation by the same GP.The measure of 'quality' was "doctor’s participation in chronic disease management programs", which is pretty coarse, so it's not surprising to not find anything using that. The 'bulk-billing' GPs essentially charge no out-of-pocket fee to their patients and only claim the subsidy from the government, so it is appropriate to see the effect on the analysis of excluding them. The extent of price discrimination is pretty striking - higher income patients may be paying 25% more for a GP consultation than lower income patients.
However, the most interesting results come a bit later in the paper, when Johar disaggregates her results based on the extent of local competition in the GP market (based on deciles of the GP-to-population ratio). She then finds:
The average fee gap in high competition areas is only $3.50, but it is more than double that in low competition areas. There is no difference in fee gap by level of local competition among GPs who selectively bulk-bill their patients, but 70% of GPs in high competition areas adopted 100% bulk-billing.Competition between doctors matters. When doctors face more local competition, they react by reducing the premium they charge higher income patients compared with those that face less local competition. Which is exactly what you would expect, since more local competition means that patients have more substitutes and relatively more elastic demand. And on top of that, GPs in high competition areas are more likely to adopt bulk-billing, and charge no additional fee to the patient.
So doctors are more likely to engage in price discrimination if they face less competition, and the difference in prices (between high-income and low-income patients) is larger when there is less competition.