Physician workload and treatment choice: the case of primary care

Primary care is a notable example of service industry where capacity-constrained suppliers face fluctuating demand levels. Unable to adjust prices, such providers may degrade quality when faced with high Little known, however, on the nature adjustments in primary context. We study how physicians trade off one key input --- their time patients other inputs, as prescriptions, lab tests and referrals. Employing detailed administrative data from eleven clinics large Israeli HMO, we use absence colleagues source exogenous variation physician workload. find no evidence that physicians' workload affects intensity which they prescribe painkillers, or refer Emergency Room, very little for an effect prescription antibiotics. do find, diagnostic inputs are complements: minute decrease (daily) average visit length causes 9 percent referrals specialists, 3.8 tests. Following recent literature, complement traditional exclusion restriction within linear model by estimating non-parametric bounds Average Treatment Effects using alternative assumptions. Such estimators rule out possibility tools substitutes economically-meaningful fashion, while still leaving broad scope those complements. Taken together, results indicate shadow cost capacity not reflected poor treatment decisions, but instead be manifested underprovision specialists fundamental aspects long-term preventive care.