Market structure friction has first-order effects on corporate payouts. The 1994 NASDAQ Manning Rule, the 1997 tick size reduction, the 2001 decimalization, and the jumpstart of NYSE algorithmic trading in 2003 dramatically increased share repurchases, whereas the 2016 Tick Size Pilot, which partially reversed these reforms, significantly reduced share repurchases. Our results provide the first unified interpretation that helps to resolve two puzzles in the payout literature. The dividend puzzle exists because previous research has overlooked the microstructure cost of executing share repurchases. Buybacks increase relative to dividends over time because market structure reforms gradually reduce the cost of repurchases.