This paper provides evidence that the arrival of an additional low-acuity patient substantially increases the wait time to start of treatment for high-acuity patients, contradicting the long-standing prior conclusion in the medical literature that the effect is \negligible.” Whereas the medical literature underestimates the effect by neglecting how delay propagates in a queuing system, this paper develops and validates a new estimation method based on queuing theory, machine learning and causal inference. Wait time information displayed to low-acuity patients provides a quasi-randomized instrumental variable. This paper shows that a low-acuity patient increases wait times for high-acuity patients through: pre-triage delay; delay of lab tests ordered for high-acuity patients; and transition delay when an ED interrupts treatment of a low-acuity patient in order to treat a high-acuity patient. Hence high-acuity patients’ wait times could be reduced by: reducing the standard deviation or mean of those transition delays, particularly in bed-changeover; providing vertical or \fast track” treatment for more low-acuity patients, especially ESI 3 patients; standardizing providers’ test-ordering for low-acuity patients; and designing wait time information systems to divert (especially when the ED is highly congested) low-acuity patients that do not need ED treatment.