The ”supplier-induced demand (SID)” hypothesis states that health suppliers, in the face of negative income shocks, may exploit their agency relationships with patients by providing excessive care. We argue SID leads providers to compensate for their losses by substituting away from drug treatment toward more heavily reimbursed dialysis treatment for chronic renal failure (CRF) patients. We test this hypothesis by examining whether a provider's earlier loss increases the volume of dialysis treatment for new CRF patients during the current period. Using Taiwan National Health Insurance Data between 1996 and 2001, we found this negative correlation after controlling for unobserved characteristics of providers (e.g., reputation) using a ”first difference” approach. On average, a provider would compensate for 5-10% of their losses with new dialysis treatment. These results holds for bofh clinics and physicians and for periods of 3, 6, or 12 months. Finally, our results show the strongest income effect occurs for non-profit hospitals supported by enterprises.