End-of-life care represents a major fiscal and social challenge facing aging economies across the globe. We study how expanding hospice services affects public spending and patient well-being near the end of life. Using 1,518 hospice facility openings across U.S. counties between 1999 and 2019, we estimate that increased access to end-of-life hospice care generates a roughly 2-to-1 fiscal return: each new facility increases hospice spending by $4.30 per Medicare beneficiary-quarter (7.1% relative to baseline) but reduces total spending by $8.80, primarily through reduced hospitalizations and post-acute care. Patients spend fewer of their final days in institutions, receive less intensive and more comfort-oriented care, and are more likely to die at home as opposed to in an institution. We find no adverse effects on pain, functional status, or mortality. Despite these benefits, hospice utilization varies widely across conditions. A simple framework relating utilization to patient severity, provider incentives, and capacity constraints suggests that relaxing supply-side barriers---particularly capacity constraints---could improve social welfare. (with Edward Kong and Maya Roy)